Vol 7 No 1 - Roger Williams University School of Law

ROGER WILLIAMS UNIVERSITYLAW REVIEWVOLUME SEVEN FALL 2001 NUMBER ONEEDITORIAL BOARDChristopher A. AndersonEditor-in-ChiefStephen P. CooneyArticles EditorChristy L. HetheringtonNotes and Comments EditorStanley F. PupeckiArticles EditorLucy H. HolmesNotes and Comments EditorDena M. CastriconeSurvey EditorSENIOR STAFFMichael J. DalyJoseph M. ProiettaSTAFFMichelle M. AlvesMark P. GagliardiCamille A. McKennaJill A. TaftMarjorie A. ConnellyJoe H. Lawson IISusan K. RodriguezBetsy WallFACULTY ADVISOREdward J. EberleADMINISTRATIVE ASSISTANTGerlinde Celonaiii

ROGER WILLIAMS UNIVERSITYLAW REVIEWVOLUME SEVEN FALL 2001 NUMBER ONETABLE OF CONTENTSSYMPOSIUM — INFORMATION AND ELECTRONICCOMMERCE LAW: COMPARATIVEPERSPECTIVESIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .David A. RiceDatabase Protection in theDigital Information Age . . . . . . . . . . . . . . . . . . . . . .Charles R. McManisDigital Property:The Ultimate Boundary? . . . . . . . . . . . .. . . . . . . . . . .Brian F. FitzgeraldWhat the Future Holds: PolicyChoices in the Global E-Marketplace . . . . . . . . . Maureen A. O’RourkeTaking UCITA on the Road:What Lessons Have We Learned? . . . . . . . . . . . . . . . . .Amelia H. BossNOTES AND COMMENTSMaking Waves In Statutory Safe Harbors:Reevaluating Internet Service Providers’Liability for Third Party Content andCopyright Infringement . . . . . . . . . . . . . . . . . . . . . . . . . .Lucy H. HolmesFBI’s Carnivore: Is the Government EatingAway Our Right of Privacy? . . . . . . . . . . . . . . . . . . . .Patricia K. HolmesPagev

ROGER WILLIAMS UNIVERSITYSCHOOL OF LAW2001 - 2002FACULTY AND ADMINISTRATIVE OFFICERSRoy J. Nirschel, M.A., Ph.D., University of Miami, University President.Bruce I. Kogan, B.A., Syracuse University; J.D., Dickinson School of Law; LL.M., GeorgetownUniversity Law Center; Professor of Law; Interim Dean.Diana Hassel, B.A., Mount Holyoke College; J.D., Rutgers, The State University of New JerseySchool of Law-Newark; Associate Dean for Academic Affairs and Associate Professorof Law.Christel L. Ertel, B.S., The Pennsylvania State University; J.D., Widener University School ofLaw; Dean of Admissions and Associate Professor of LawGail I. Winson, B.A., Moravian College; M.S., Drexel University; J.D., University of Florida;Director of Law Library and Associate Professor of Law.Chelsie Horne, B.S., M.B.A., C.M.P., Johnson & Wales University; Coordinator of Alumni,Programs and Events.Laurie Barron, B.A., Yale University; M.S.W., New York University School of Social Work;J.D., New York University School of Law; Director of the Feinstein Institute for LegalService.Linda J. Vieira, B.S., Roger Williams University; Director of Student Finance and Records.Nancy Waggner, B.S., Northeastern University; J.D., Suffolk University Law School; Director ofCareer Services.Barbara L. Bernier, B.A., State University of New York at Plattsburgh; M.S.W., AdelphiUniversity; J.D., Howard University School of Law; LL.M., Temple University Schoolof Law; Associate Professor of Law.Carl T. Bogus, A.B., J.D., Syracuse University; Associate Professor of Law.Esther F. Clark, B.A., Temple University; J.D., Rutgers, The State University of New JerseySchool of Law-Camden; Distinguished Visiting Professor of Law.Debra R. Cohen, A.B., Brown University; J.D., Emory University School of Law; VisitingAssociate Professor of Law.Edward J. Eberle, B.A., Columbia College; J.D., Northwestern University School of Law;Professor of Law.Jonathan M. Gutoff, A.B., Brown University; J.D., University of Chicago School of Law;Associate Professor of Law.Matthew P. Harrington, B. Th., McGill University; J.D., Boston University School of Law;LL.M., S.J.D., University of Pennsylvania Law School; Professor of Lawand Director of the Marine Affairs Institute.Andrew Horwitz, B.A., Haverford College; J.D., New York University School of Law;Associate Professor of Law and Director of the Criminal Defense Clinic.Robert B. Kent, B.A., Harvard University; LL.B., Boston University School of Law;Distinguished Visiting Professor of Law.Peter C. Kostant, B.A., Yale College; M.A., Washington University of St. Louis; J.D.,Fordham University School of Law; Associate Professor of Law.John C. Kunich, B.S., M.S., University of Illinois; J.D., Harvard University Law School;LL.M., George Washington University School of Law; Associate Professor of Law.Peter S. Margulies, B.A., Colgate University; J.D., Columbia University School of Law;Professor of Law and Director of the Disability Law Clinic.Lisa T. McElroy, B.A., Dartmouth College; J.D., Harvard Law School; Professor of LegalWriting.Colleen P. Murphy, B.A., University of Virginia; J.D., Yale Law School; Professor of Law.David A. Rice, B.B.A., University of Wisconsin; LL.B., Columbia University School of Law;Professor of Law.Harvey Rishikof, B.A., McGill University; M.A., Brandeis University; J.D., New YorkUniversity School of Law; Professor of Law.Larry J. Ritchie, B.A., J.D., University of South Carolina; LL.M., Georgetown University LawCenter; Professor of Law.vii

Emily J. Sack, B.A., Swarthmore College; M.A., M.Phil, Columbia University; J.D., NewYork University School of Law; Associate Professor of Law.Ellen M. Saideman, B.A., Barnard College; J.D., Columbia University School of Law;Professor of Legal Writing.Anthony J. Santoro, B.A., Boston College; J.D., LL.M, Georgetown University Law Center;President Emeritus of Roger Williams University and Senior Professor of Law.B. Mitchell Simpson III, A.B., Colgate University; M.A., M.A.L.D., Ph.D., Tufts University;LL.B., University of Pennsylvania Law School; Professor of Legal Writing.Louise E. Teitz, B.A., Yale College; J.D., Southern Methodist University School of Law;Associate Professor of Law.Michael J. Yelnosky, B.S., University of Vermont; J.D., University of Pennsylvania LawSchool; Associate Professor of Law.David M. Zlotnick, B.A., State University of New York at Binghamton; J.D., Harvard LawSchool; Associate Professor of Law.viii

Identification StatementThe Roger Williams University Law Review is published twiceannually by the students of the Roger Williams University Law Review,10 Metacom Avenue, Bristol, Rhode Island 02809; gbc@rwulaw.rwu.edu(electronic mail); http://www.rwu.edu/law/lawrev.html (internet address).Subscription rates: $28.50 per year in the United States; $33.00 per yearin foreign countries. Single issues: $15.00 each. Address allsubscription requests and other business communications to the attentionof the Administrative Assistant, Roger Williams University Law Review.If a subscription is to be discontinued at the expiration of the currentvolume, then notice to that effect should be sent to the AdministrativeAssistant; otherwise, subscriptions will be automatically renewed. Onemonth’s notice is necessary to effect change.Citations generally conform to The Bluebook: A Uniform System ofCitation (17th ed. 2000). Copyright © 2000 by The Columbia LawReview Association, the Harvard Law Review Association, theUniversity of Pennsylvania Law Review and The Yale Law Journal.______________________________________________________Submission of Articles and LettersThe Roger Williams University Law Review accepts unsolicitedmanuscripts and letters. All manuscripts must be typewritten, doublespacedand on 8½ x 11" non-bond paper; footnotes should conform toThe Bluebook: A Uniform System of Citation.The Roger Williams University Law Review reserves the right to editall manuscripts and letters. However, it is requested that such permissionbe given to the Roger Williams University Law Review.No manuscripts can be returned unless accompanied by a stamped,self-addressed envelope of appropriate size. Letters cannot be returned.Acknowledgment of receipt of manuscripts will be made inapproximately three weeks.Copyright is reserved by the Roger Williams University Law Reviewto all material published by the Roger Williams University Law Review,absent express agreement to the contrary.ix

Symposium – Information andElectronic Commerce Law:Comparative PerspectivesIntroductionDavid A. Rice*Information technologies and electronic commerce presentdomestic and international law challenges on many dimensions.The Internet, World Wide Web and telecommunicationstechnologies facilitate information dissemination, access and usewithout regard for national boundaries. Technological innovationsinfuse commercial markets and transactions with changes thatpresent new prospects, risks and legal issues. Software anddatabase protection issues challenge traditional intellectualproperty paradigms. Personal data collection, freedom ofinformation access and freedom of communication pose newquestions within domestic and international political realms.Other critical issues arise in the course of adapting privatecontract law to electronic contracting and from the extension ofintellectual property licensing notions to the distribution of digitalinformation products. Standard form contract regulation of copyuse transfer and restriction raise anew issues that have longvexedthe law in what now seem like more mundane sale of goods* Professor, Roger Williams University School of Law. B.B.A., University ofWisconsin; LL.B., Columbia University School of Law. Professor Rice was theorganizer of this symposium.237

contract exclusion or disclaimer of warranties and contractualreduction of remedies for breach. Use of standard forms andterms now pose issues about the efficacy of contract to negate thefirst sale doctrine of copyright law as well as the right under statetrade secret law and the fair use doctrine of copyright law toreverse engineer computer program object code.Moving further into the contractual realm, digital informationtechnology has introduced on-line business-to-business, consumerto-business,citizen-to-government, and government-to-citizenelectronic transactions. “Digital signature” and electronicdocument authenticity issues are the focus of state and nationallawmaking, European Community legislation and internationalmodel laws and treaties.Substantial economic interests are at stake. Making themarket safe for distribution of digital content and works isessential to development of new and evolving industries. It iscritical to the realization of benefits offered by greater access toinformation content, functional works and entertainment productsmade possible by digital and communications technologies.Equally, legal certainty and harmonization concerningauthenticity, or risk of inauthenticity, of digital signatures anddocument contents are essential to stability and development ofelectronic contracting.The tensions between private and public interestscharacterize all debate on these issues, and differing perspectivesexist both within and among diverse cultures and differing legalsystems. Recognition of this was the inspiration for myorganization of the April 20, 2001 Scholarly SymposiumInformation and Electronic Commerce Law: ComparativePerspectives and the informal speakers’ and moderators’Roundtable discussion the following morning. The event featuredeight leading United States, Portuguese and Australian lawscholars. Four prominent and highly regarded information andelectronic commerce attorneys moderated the Symposiumsessions,1 and offered substantive comments when introducingtheir session and led question and answer discussions following1. The session moderators were Jonathan Band, Washington, D.C., office ofMorrison & Foerster L.L.P.; Wayne D. Bennett of Bingham Dana LLP in Boston,Massachusetts; Jerry Cohen from the Boston, Massachusetts, office of Tillinghast LichtPerkins Cohen & Smith L.L.P.; and Marc S. Friedman of the Roseland, New Jersey,office of Goodwin Proctor LLP and President of the Computer Law Association.

speaker presentations.The Symposium was co-sponsored by the Roger WilliamsUniversity Law Review and the Roger Williams University Schoolof Law Portuguese-American Comparative Center. TheRoundtable, which I moderated, was sponsored by the newlyestablished Center. The informality of the latter provided a rareopportunity for a direct, sustained and lively exchange of viewsamong leading scholars and lawyers brought together from quitedistant places and from diverse cultures and legal systems.Together, the Symposium and Roundtable experiences happilyreminded their participants that diversity of human experience,culture and critical thought persists within our networkedinformation society even as nations seek increasing harmonizationof intellectual property and contract law important to economicactivity, research and discourse in a network environment.Symposium presentations were made in four separatesessions: Digital Electronic Database Protection, Technology inElectronic Commerce, Electronic Commerce Transactions, andInformation and Electronic Commerce Horizons. Four of theimportant papers presented at the Symposium are published inthis issue of the Law Review. Unexpected circumstances made itimpossible for two of the speakers to submit expanded versions oftheir presentation, but the program itself was nonetheless greatlyenriched by the formal remarks of Professor Peter A. Jaszi and Dr.Antõnio de Macedo Vitorino and their engagement in discussionand participation in the Roundtable. Professor Jaszi2 addressedthe extension of digital content provider control over access to anduse of information content under the Digital MillenniumCopyright Act and digital content use licensing. Dr. Vitorino3spoke on doing business over the Internet. He demonstrated theimportance of businesses and legal counsel being aware of andconversant with contract, intellectual property, data protection2. Peter A. Jaszi is Professor of Law at American University Washington Collegeof Law. Professor Jaszi is the Director of the Glusko-Samuelson Intellectual PropertyLaw Clinic, a project of the American University Washington College of Law Programon Intellectual Property and the Public Interest.3. Antõnio de Macedo Vitorino is a Senior Partner in the Lisbon, Portugal lawfirm Macedo Vitorino & Associates and Assistant Professor, University of LisbonFaculty of Law. Dr. Vitorino, in his law practice, specializes in competition, securities,intellectual property and commercial transactions law with particular emphasis onemerging issues concerning the digital electronic environment. He is a FoundingMember and Director of the Portuguese Intellectual Property Law Association.

and other laws of countries in which users and customers arelocated rather than assuming that only the familiar law of acompany’s own jurisdiction governs its transactions in theInternet environment.Professor Brian Fitzgerald, Dean of Southern CrossUniversity School of Law and Justice in Lismore, Australia,authored and presented the article titled Digital Property: TheUltimate Boundary? Dean Fitzgerald is a leading scholar andauthority on Australian intellectual property and Internet law,and often participates in international dialogue as a speaker andauthor of articles published in international journals. Hispresentation and article weigh the role and limits of law intransforming information and information products into economicvalue, wealth and industries in the digital environment. DeanFitzgerald seeks to harmonize public and private interests ininformation, and to address the challenge of regulating in such away as to promote and secure a return to new investment withoutsacrifice of access to and use of knowledge by private individuals,researchers and other businesses. He brought to the Symposiumand this issue of the Law Review the perspective of a scholar froma nation that is a very significant but not a dominant producerand exporter of digital works.Charles R. McManis, Professor of law at WashingtonUniversity School of Law in St. Louis, is a leading scholar andpublic voice on copyright, contract and protection of computersoftware and other digital information products. Although he haspublished important works on United States law, he also hastaught and studied intellectual property law in Asia and madeinternational and comparative aspects of intellectual property lawa particular focus of his scholarship. Professor McManis’spresentation and article, Database Protection in the DigitalInformation Age, brings to his topic a depth of study and a graspof policy and legal issues that has made his previous writing onnational and international intellectual property law issuesessential reading for contemporary scholars and policy makers.Professor McManis’s particular concern in his article is with thetwo principal, and competing, proposals for United Statesdatabase protection legislation. He informs his readers with apreliminary examination of the European Community’s creation ofsui generis database extraction and reutilization rights as anadjunct to the very limited database protection available under

copyright law. Professor McManis then thoughtfully andinsightfully addresses whether and to what extent United Stateslegislation is required in order to satisfy the internationalreciprocity provision in the European Community DatabaseDirective, market considerations and legal considerations in amanner that equally speaks to scholars, lawyers and public policydecision makers.Professor Amelia H. Boss, the Charles Klein Professor of Lawand Government at Temple University, presented and has writtenTaking UCITA on the Road: What Lessons Have We Learned?.Professor Boss is a member of the Council of the American LawInstitute (ALI), chairs the Business Law Section of the AmericanBar Association and serves as a United States Delegate to UnitedNations Commission on International Trade Law (UNCITRAL) onelectronic commerce issues. She was an ALI member of theDrafting Committee for a Proposed Article 2B of the UniformCommercial Code (now the Uniform Computer InformationTransactions Act, or UCITA) and is an ALI member of theDrafting Committee to revise Articles 1, 2 and 2A of the UniformCommercial Code. It should be clear from this impressive, yetselective, review of Professor Boss’s experience that there is noperson better qualified to address the need for new contract lawrules, the approach and particulars of the UCITA, and howUCITA compares with other modern commercial contract lawstatutes. The dimensions in which she does so in her articleinclude contract law theory, choice between general andparticularistic statutory expression of rules, soundness of policychoices reflected in certain rules and suitability of the uniform lawfor export as a model for digital information product and servicestransactions legislation.Among scholars with whom I have had occasion to disagree onimportant information technology copyright and contract lawissues over the last half-dozen years, Professor Maureen AnneO’Rourke of Boston University School of Law is one for whom Ihave greatest regard. Her critical, creative, insightful andrigorous thought challenges others to likewise extend themselvesas participants in critical thinking and dialogue rather than adogmatic contest. Although a relatively new member of theacademic cadre, she is a recipient of Boston University’s prizedMetcalf Award for outstanding teaching and her writings arewidely regarded as “must” reading. Professor O’Rourke closed the

Symposium and writes on this issue in What the Future Holds:Policy Choices in the Global E-Marketplace. What greaterincentive can I give the reader to spend time with her article thanto say that Professor O’Rourke’s presentation held the Symposiumaudience in place even though it was the Friday afternoon closingpresentation in a long, full-day program?As experienced and as viewed in retrospect, the full daySymposium and half-day Roundtable programs were remarkablefor the ease in which academics and lawyer-scholars from verydifferent legal and cultural backgrounds found common purposeand genuine camaraderie, and an enjoyment in sharing that withRoger Williams University law students and faculty. As theorganizer and convener of the programs, I wish to express mygratitude to professional colleagues and personal friends whoaccepted my invitation to participate in the programs asmoderators and as speakers. Writing this introduction alsoprovides me a forum to express my gratitude for a personally quitespecial benefit and satisfaction, the privilege of meeting, andbecoming friends as well as professional colleagues, of a mostdistinguished group of scholars from abroad: Professor DoutourJosé de Oliveira Ascensão, Dean Brian Fitzgerald, Dr. ManualLopes Rocha and Dr. Antõnio de Macedo Vitorino.Finally, though not least, I speak for all who participated inthe Symposium and Roundtable in closing this introduction byexpressing gratitude to others who made the event andpublication of this Symposium Issue notable in the young life ofthis law school. We are especially grateful to the Luso-AmericanDevelopment Foundation and its President, Dr. Rui Chancerellede Machete for program and financial support, the Law Reviewand its members for their work in editing and publishing thepapers in this Symposium Issue, and Dr. Manuel Duarte deOliveira of the Luso-American Development Foundation for hisliaison assistance in Portugal. Most especially, I speak for all inrecognizing my colleague, and at time the Dean, Professor HarveyRishikof for his encouragement, unstinting support, and spiritedhospitality that made our distinguished guests’ visits to theUniversity and the School of Law so congenial and memorable.

Database Protection in the DigitalInformation AgeCharles R. McManis *It would be an atrocious doctrine to hold that dispatches,the result of the diligence and expenditure of one man,could with impunity be pilfered and published byanother . . . . The mere fact that a certain class ofinformation is open to all that seek it, is no answer to aclaim to a right of property in such information made by aperson who, at his own expense and by his own labor, hascollected it.1INTRODUCTIONAs industries in the global market increasingly come to relyon electronic compilations of data,2 calls for new, sui generis formsof legal protection for databases have grown apace. Yet, as thequotation above illustrates, the issue of data protection is notnew—nor is the world without precedent in determining how theissue of data protection ought to be resolved.*Professor of Law, Washington University in St. Louis; B.A., 1964, Birmingham-Southern College; M.A. (Philosophy); J.D., 1972, Duke University. The authorgratefully acknowledges the research assistance of Christopher R. Alder, J.D., 2001,Washington University in St. Louis, who also provided initial drafts of portions of thisarticle. The author is likewise indebted to Washington University law librarians, HylaBondareff and Katrina Stierholz, for their research assistance at several critical pointsin the preparation of this article. Finally, the author gratefully acknowledges theinsightful comments and suggestions of Justin Hughes, Visiting Professor of Law,UCLA School of Law, who graciously read and commented up an initial draft of thisarticle.1. Kiernan v. Manhattan Quotation Telegraph Co., 50 How. Pr. 194, 14 C.O.Bull. 1493 (N.Y. Sup. Ct. 1876).2. See Robert M. O’Neil, Campus Database Issues, 27 J.C. & U.L. 109 (2000).237

Indeed, the quotation contains the seeds of two possibleapproaches to data protection, and therein lies the problem.Should the law focus more on the data pilferer’s wrong or on thedata collector’s claim to a property right in the data? In otherwords, is the legal solution to the problem of database protectionto be found in the law of unfair competition or the law ofintellectual property?While the United States, which currently dominates the worlddatabase industry,3 remains mired in debate over which approachto take, the European Union has aggressively opted to create anew intellectual property right in data. In 1996, after nearly eightyears of discussions over how to improve “a very great imbalancein the level of investment in the database sector both as betweenthe Member States and between the Community, and the world’slargest database-producing third countries”4 (the latter being codewords for the United States), the European Union promulgated itsDirective on the Legal Protection of Databases (EU DatabaseDirective). The Directive not only mandated the harmonization ofEuropean copyright law with respect to the protection ofcompilations of data, but also obliged its members to create a new,sui generis form of intellectual property protection for the contentsof the database itself, irrespective of its eligibility for copyright orother legal protection. Article 7 of the Directive requires MemberStates to provide “a right for the maker of a database which showsthat there has been qualitatively and/or quantitatively asubstantial investment in either the obtaining, verification orpresentation of the contents to prevent extraction and/or reutilizationof the whole or of a substantial part, evaluatedqualitatively and/or quantitatively, of the contents of thatdatabase.”5 Coupled with this controversial new intellectual3. See The Collections of Information Antipiracy Act and The Vessel Hull DesignProtection Act: Hearing on H.R. 2652 and H.R. 2696 Before the Subcomm. on Courtsand Intellectual Prop. of the House Comm. on the Judiciary, 105th Cong. (1997)[hereinafter Kirk Statement] (statement of Michael K. Kirk, Executive Director,American Intellectual Property Law Association) (stating that according to the GaleDirectory of Databases, “the U.S. produces 64.2% of all the world’s databases.Compared to the 6000+ databases produced in the United States, only nine othercountries produce more than 100 databases (Australia, Canada, England, France,Germany, Italy, Japan, Netherlands, and Spain).”), available athttp://www.house.gov/judiciary/106-kirk.htm.4. Council Directive 96/9/EC of 11 March 1996, on the Legal Protection ofDatabases, 1996 O.J. (L 77) 20, recital 11 [hereinafter EU Database Directive].5. Id. at art. 7.

property right, the Database Directive contains an equallycontroversial reciprocity provision, which specifies that this rightwill apply to databases whose makers are nationals or habitualresidents of third countries “only if such third countries offercomparable protection to databases produced by nationals of aMember State or persons who have their habitual residence in theterritory of the Community.”6 In other words, unless the UnitedStates enacts comparable protection for European databases, U.S.database producers other than those who also establishthemselves in an EU member state will not be entitled to this suigeneris protection.Not surprisingly, within weeks of the promulgation of the EUDatabase Directive, the first U.S. bill aimed at creating a similarsui generis right in data was proposed in the United States Houseof Representatives.7 Yet today, after five years of pitched debate,most recently over two bills embodying the two competingapproaches to the problem of data piracy, all the United Statesseems to have achieved is a legislative stalemate on the issue.In the now-expired 106th Congress, committees of the UnitedStates House of Representatives considered two bills—H.R. 354(the “Collections of Information Antipiracy Act”)8 and H.R. 1858(the “Consumer and Investor Access to Information Act”).9 H.R.354 would effectively create a sui generis right, somewhatanalogous to that mandated by the EU Database Directive, byprohibiting the extraction or making available to others of all or asubstantial part of a collection of information gathered, organizedor maintained by another person through the investment ofsubstantial monetary or other resources, so as to cause materialharm to the primary or a related market of that other person or asuccessor in interest.10 Though labeled an “antipiracy” act andcontaining features, such as the “material harm” requirement,6. Id. at recital 56.7. See infra note 149 and accompanying text.8. H.R. 354, 106th Cong. (1999). Actually, there are two versions of this bill, theoriginal bill as introduced in the House, dated Jan. 19, 1999, and a revised bill asreported in the House (Union Calendar No. 212), dated Oct. 8, 1999. Unless otherwisenoted, all references in this article will be to the later version of the bill.9. H.R. 1858, 106th Cong. (1999). There are likewise two versions of this bill,the original bill as introduced in the House, dated May 19, 1999, and a revised bill asreported in the House (Union Calendar No. 213), dated Oct. 8, 1999. All referencesherein are to the later version of the bill.10. H.R. 354 § 1402.

suggestive of a bill designed merely to prevent unfair competition,H.R. 354 nevertheless extends protection to both the primarymarket of a database producer and to any related market, anddefines a “related market” sufficiently broadly that a databaseproducer would be able control a wide range of non-competitiveuses of data and reserve those potential markets for itself.11By contrast, H.R. 1858 would adopt a much narrowerapproach to data piracy by prohibiting (1) the use of any means orinstrumentality of interstate or foreign commerce orcommunication to sell or distribute to the public a database that isa duplicate of another database and is sold or distributed incompetition with the original database and (2) themisappropriation of real-time market information.12 Indeed, H.R.1858 arguably does little more than federally codify the commonlawtort of “hot news” misappropriation,13 as it might apply todatabases used in interstate and foreign commerce, and providethe Federal Trade Commission with enforcement authority. Likethe common-law tort of “hot news” misappropriation, H.R. 1858responds straightforwardly to the problem of data piracy byprohibiting the wholesale duplication of databases for competitivepurposes or the misappropriation of real-time market information,while at the same time ensuring that transformative uses can bemade of factual compilations.14 By contrast, H.R. 354, like the EUDatabase Directive, will provide database producers withsubstantial control over what transformative uses can be made ofdatabases, thus potentially undermining competitive conditions inthe database industry and discouraging innovative uses of thevery databases whose production it is designed to stimulate.Apparently aware of precisely that possibility, the EU specified inArticle 16(3) of the Database Directive that by the end of 2001,and every three years thereafter, a report is to be prepared toverify whether the application of the sui generis right “has led toabuse of a dominant position or other interference with free11. Id. § 1401(4). For a detailed discussion of the bill’s definition of a “relatedmarket,” see infra notes 159-62 and accompanying text.12. H.R. 1858 §§ 102, 201.13. For a discussion of the genesis of this common law tort, see infra notes 34-40and accompanying text.14. See generally Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 579 (1994)(discussing the importance of being able to make transformative uses of thecopyrighted works of others).

competition which would justify appropriate measures beingtaken, including the establishment of non-voluntary licensingarrangements.”15The emergence of two competing bills in the United StatesCongress suggests that the reciprocity provision of the EUDatabase Directive is skewing the U.S. debate over databaseprotection, just as it was intended to do. Nor is the United Statesin much of a position to protest this use of a reciprocity provisionthat foists off a new form of intellectual property protection on anunwilling world, as the United States itself used precisely thesame tactic in 1984 to induce the rest of the world to create suigeneris protection for semiconductor chip designs.16 Perhaps thebest one can hope for is that the legislative stalemate in Congressmight continue until either the economic consequences of the EUDatabase Directive or the legality of its reciprocity provision, orboth, can be determined. However, U.S. congressional leaders aresaid to have already decided that a database bill will beintroduced in the current session of Congress, and negotiationswith the committee staffs responsible for H.R. 354 and H.R. 1858have begun.17 The question is what sort of database bill the15. EU Database Directive, supra note 4, at art. 16(3). It appears that the EU’sconcerns were not unfounded. See Commission Decision 01/165/EC of 3 July 2001,Case COMP D3/38.044—NDC Health/IMS Health: Interim measures, 2002 O.J. (L 59)18 (granting interim measures requiring sales data company to license its copyrighteddatabase structure to its competitors based on a finding of abuse of its dominantposition). However, this decision was later suspended by the President of the Court ofFirst Instance pending further examination upon appeal; see Order T-184/01 R 2 (Oct.26, 2001), available at http://curia.eu.int/jurisp/cgi-bin/form.pl?lang=en; see also ECCompetition Policy Newsletter, Feb., 2002, at 61, available athttp://europa.eu.int/comm/competition/publications/cpn/cpn2002_1.pdf.16. See Charles R. McManis, Taking TRIPS on the Information Superhighway:International Intellectual Property Protection and Emerging Computer Technology, 41Vill. L. Rev. 207, 258-59 (1996) [hereinafter McManis, Taking TRIPS]; Charles R.McManis, International Protections for Semiconductor Chip Designs and the Standardof Judicial Review of Presidential Proclamations Issued Pursuant to the SemiconductorChip Protection Act of 1984, 22 Geo. Wash. J. Int’l L. & Econ. 331, 338-39 (1988).17. See Robert MacMillan, House Committees Try To Thaw Database Bill Freeze,Newsbytes (Mar. 29, 2001) (reporting the reopening of discussions to try to fashion acommon database intellectual property bill out of the two conflicting bills), athttp://www.newsbytes.com/news/01/163826.html; Washington Affairs Office, Am. Ass’nof Law Libraries, House Legislation to Protect Databases (July 2001) (reporting on 8weekly meetings, co-hosted by House Judiciary Committee Chairman JamesSensenbrenner and House Energy & Commerce Committee Chairman Billy Tauzin,between proponents and opponents of database legislation), available athttp://www.ll.Georgetown.edu/aallwash/ib0720012.html.

United States ought to adopt in light of the challenge posed by theEU Directive.Accordingly, Part I of this Article will examine how theUnited States law of copyrights and unfair competition, as well asthe emerging law of electronic contracts, currently protectscompilations of data, and what additional legislative protection isbeing proposed. Part II will examine how the changing face of thedatabase industry in the new digital age has enhanced both thevalue and the vulnerability of investments in electroniccompilations of data, thus precipitating the current spate oflegislative proposals for sui generis database protection. Part IIIwill examine the particular conditions and motivations that led tothe promulgation of the EU Database Directive, and describe thesui generis protection mandated therein. Finally, Part IV willcompare the two most recent United States legislative proposals,both with each other and with the EU Database Directive,concluding with some thoughts about what the United Statesgovernment should do in light of the dilemma it faces inresponding to the challenge posed by the EU Database Directive.I. PROTECTION OF COMPILATIONS OF DATA: CURRENT U.S. LAW ANDPROPOSED LEGISLATIONIn the United States, calls for sui generis database protectionbegan to mount in the wake of the United States Supreme Court’s1991 ruling in Feist Publications, Inc. v. Rural Telephone ServiceCo.,18 which held that white pages of a telephone directory werenot sufficiently original and creative to warrant copyrightprotection.19 In Feist, the Court reiterated its long-standinginsistence “that the fact/expression dichotomy limits severely thescope of [copyright] protection in fact-based works.”20 Specifically,it held that:Facts, whether alone or as part of a compilation, are notoriginal and therefore may not be copyrighted. A factualcompilation is eligible for copyright if it features anoriginal selection or arrangement of facts, but thecopyright is limited to the particular selection or18. 499 U.S. 340 (1991).19. Id. at 364.20. Id. at 350.

arrangement.21In so holding, the Court repudiated a line of lower court cases thatheld that compilations of data were copyrightable per se, merelybecause they were the product of the “sweat of the brow” of theirproducers.22For almost a century, U.S. copyright protection had seemed toderive from two not altogether consistent theoretical models. Onthe one hand, the courts insisted that the sine qua non ofcopyright protection was the requirement of “originality,” whichwas defined to mean not only that the work was an independentcreation of the author (as opposed to being copied from otherworks), but also that it possessed at least some minimal degree ofcreativity.23 On the other hand, a competing line of lower courtdecisions involving compilations of data also embraced the “sweatof the brow” or “industrious collection” concept, which viewedcopyright protection as a reward for the hard work that went intocompiling facts.24The “sweat of the brow” doctrine, however, was seriouslyundermined by the Copyright Act of 197625 and then definitivelyput to rest by the Supreme Court in Feist. Section 103 of the 1976Copyright Act explicitly included compilations within the subjectmatter of copyright, but section 101 defined a “compilation” as a“work formed by the collection and assembling of preexistingmaterials or of data that are selected, coordinated, or arranged insuch a way that the resulting work as a whole constitutes anoriginal work of authorship.”26 Notwithstanding this language,some lower courts continued to reward “sweat of the brow” byextending copyright protection to mere facts,27 thus stimulating21. Id.22. Id. at 352-61.23. See The Trade-Mark Cases, 100 U.S. 82 (1879); Burrow-Giles Lithograph Co.,v. Sarony, 111 U.S. 53 (1884).24. See, e.g., Jeweler’s Circular Publ’g Co. v. Keystone Publ’g Co., 281 F. 83 (2dCir. 1922); Leon v. Pac. Tel. & Tel. Co., 91 F.2d 484 (9th Cir. 1937).25. 17 U.S.C. §§ 101-1332 (1994).26. Id. § 101.27. See, e.g., West Publ’g. Co. v. Mead Data Cent., Inc., 799 F.2d 1219 (8th Cir.1986) (holding that Mead Data Central’s proposal to introduce “star pagination”showing West’s page numbers in its own LEXIS database service would constitutecopyright infringement).27. See, e.g., West Publ’g. Co. v. Mead Data Cent., Inc., 799 F.2d 1219 (8th Cir.1986) (holding that Mead Data Central’s proposal to introduce “star pagination”showing West’s page numbers in its own LEXIS database service would constitute

the Supreme Court to reiterate in Feist that compilations arecopyrightable only if and to the extent that their selection orarrangement of data is original.28 As if to preempt any subsequentlegislative modification of its holding, the Court emphasized that“originality” of expression is not merely statutory but aconstitutional requirement for United States copyrightprotection.29The Feist decision has not only brought the lower courts in theUnited States to heel,30 but also seems to have gained a measureof international recognition,31 and the statutory language onwhich it was based subsequently found its way into the text of theGATT/WTO Agreement on Trade Related Aspects of IntellectualProperty Rights, commonly known as the TRIPS Agreement.32Article 10(2) of TRIPS states that “compilations of data or othermaterial, whether in machine readable or other form, which byreason of the selection or arrangement of their contents constituteintellectual creations shall be protected as such.”33 Far fromquieting the debate over the protection of databases, however, theFeist decision and Article 10 of TRIPS have merely served totransform the terms of the debate, both in the United States andelsewhere.Frustrated with the Feist decision and the limited applicationof copyright laws to factual compilations, some United Statesdatabase owners have looked to state misappropriation law tocopyright infringement).28. See Feist, 499 U.S. at 350.29. Id. at 358-62.30. See Matthew Bender & Co. v. West Publ’g. Co., 158 F.3d 693 (2d Cir. 1998)(holding pagination not copyrightable); Warren Publ’g., Inc. v. Microdos Data Corp.,115 F.3d 1509 (11th Cir. 1997), cert. denied, 522 U.S. 936 (1997) (holding that acomprehensive compilation of cable system providers was not sufficiently original andwas not protected against wholesale copying).31. See, e.g., Tele-Direct (Publications) Inc. v. Am. Bus. Info., Inc., 76 C.P.R. (3d)296, 1997 CPR Lexis 1276 (Fed. Ct. of App., Quebec, 1997) (citing Feist and other U.S.lower court decisions as “useful guides” in a case holding a compilation of informationcontained in Yellow Pages directories to be uncopyrightable under Canadian law).31. See, e.g., Tele-Direct (Publications) Inc. v. Am. Bus. Info., Inc., 76 C.P.R. (3d)296, 1997 CPR Lexis 1276 (Fed. Ct. of App., Quebec, 1997) (citing Feist and other U.S.lower court decisions as “useful guides” in a case holding a compilation of informationcontained in Yellow Pages directories to be uncopyrightable under Canadian law).32. Agreement on Trade-Related Aspects of Intellectual Property Rights, Apr. 15,1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C,Legal Instruments – Results of the Uruguay Round vol. 31, 33 I.L.M. 81 (1994).33. Id. at art. 10(2).

protect their investments. Ironically, the common-law concept of“hot news” misappropriation traces its roots to another seminalUnited States Supreme Court case, International News Service v.Associated Press,34 decided some seventy years prior to Feist. TheInternational News case is a relic from an earlier era in whichfederal courts created their own “federal common law” forapplication in diversity of citizenship cases. The Court’s reasoningin International News is reminiscent of the colorful New YorkSupreme Court quotation appearing at the outset of this article,denouncing the “atrocious doctrine . . . [that would] hold thatdispatches, the result of the diligence and expenditure of one man,could with impunity be pilfered and published by another.”35 InInternational News, the Court held that a news service had a“quasi-property” interest in the news that it had gone to theexpense of collecting,36 and could thus prevent, on a theory ofunfair competition, a rival news service from appropriating thatnews for a competitive purpose even after the news had becomepublic.37Although International News, as a part of the federal commonlaw, was effectively overruled by the Supreme Court’s laterdecision in Erie Railroad Co. v. Tompkins,38 which held thatfederal courts are to apply state common law in diversity cases,39its underlying principles have survived in state case law, andindeed, have been applied even in the absence of any competitionbetween the parties, thus coming perilously close to creating defacto (or quasi-) property rights in data.40 In Feist itself, the34. 248 U.S. 215 (1918); see also Bd. of Trade v. Christie Grain & Stock Co., 198U.S. 236 (1905).35. Kiernan, 50 How. Pr. at 194.36. Int’l News, 248 U.S. at 236. It is important to note, however, that the Court’scharacterization of the news as “quasi property” was primarily designed to sustain theCourt’s equity jurisdiction over the controversy, as a court of equity generally “concernsitself only in the protection of property rights,” but “treats any civil right of a pecuniarynature as a property right.” Id. (citations omitted).37. Id.38. 304 U.S. 64 (1938).39. Id. at 78-80. Pockets of federal common law may nevertheless remain, yet are“few and restricted.” See Atherton v. FDIC, 519 U.S. 213, 218 (1997) (citing O’Melveny& Myers v. FDIC, 512 U.S. 79, 87 (1994) (quoting Wheeldin v. Wheeler, 373 U.S. 647,651 (1963))).40. See, e.g., Bd. of Trade of the City of Chicago v. Dow Jones & Co., 456 N.E. 2d84 (Ill. 1983) (holding that offering a commodity futures contract based on the DowJones Industrial Average without the consent of Dow Jones & Co. wouldmisappropriate a valuable business asset of Dow Jones & Co., even though the latter

United States Supreme Court quoted with seeming approval arespected commentator’s observation that protection of facts andideas “may in certain circumstances be available under a theory ofunfair competition.”41 More recently, in National BasketballAssociation v. Motorola, Inc.,42 the Court of Appeals for the SecondCircuit ruled that although sports scores were mere facts, thuslacking sufficient originality to qualify for federal copyrightprotection, they might nevertheless be protected againstmisappropriation as “hot news” despite their lack of originality.43Significantly, however, the court went on to hold that “hot news”misappropriation claims would survive the federal preemptionprovisions of U.S. copyright law only when the plaintiff and thedefendant are in direct competition and the appropriation, if notprohibited, would threaten the very existence of the product orservice provided by the plaintiff.44 Because this had not beenshown, the court ordered the NBA’s claim for misappropriationdismissed.45Database owners have also sought to avail themselves ofvarious forms of federal and state legislative protection. Forexample, the Electronic Communications Privacy Act46 protectselectronic communications from interception and provides civilremedies and criminal penalties for violations of the Act.47 TheAct also prohibits intentionally accessing without authorization afacility through which an electronic communication service isprovided and thereby obtaining access to an electronicdid not deal in commodity futures contracts and had no plans to do so).41. Feist, 499 U.S. at 354 (quoting Melville B. Nimmer & David Nimmer, Nimmeron Copyright § 3.04, at 3-23 (footnote omitted)).42. 105 F.3d 841 (2d Cir. 1997).43. Id. at 852-53; but see Alcatel USA, Inc. v. DGI Technologies, Inc., 166 F.3d772, 786-89 (5th Cir. 1999) (finding federal preemption).44. Nat’l Basketball Ass’n, 105 F.3d at 853.44. Nat’l Basketball Ass’n, 105 F.3d at 853.45. Id. at 853-54.46. 18 U.S.C. §§ 2510-22, 2701-11 (1994).47. An “electronic communication” is defined as “any transfer of signs, signals,writing, images, sounds, data, or intelligence of any nature transmitted in whole or inpart by a wire, radio, electromagnetic, photoelectronic or photooptical system thataffects interstate or foreign commerce . . . .” 18 U.S.C. § 2510(12).47. An “electronic communication” is defined as “any transfer of signs, signals,writing, images, sounds, data, or intelligence of any nature transmitted in whole or inpart by a wire, radio, electromagnetic, photoelectronic or photooptical system thataffects interstate or foreign commerce . . . .” 18 U.S.C. § 2510(12).

communication while it is in electronic storage in such a system.48Database developers have also sought protection under theLanham Act. Claims for false designation of origin under section43(a), passing off, and trademark dilution under section 43(c) arepotential causes of action for database developers againstpirates.49 When the source and reliability of data is significant tothe intended users, these Lanham Act claims may proveefficacious.With the enactment of the Digital Millennium Copyright Act(DMCA) of 1999,50 yet another form of protection becameavailable to database producers. The DMCA implements theWorld Intellectual Property Organization Copyright Treaty51 byprohibiting circumvention of copyright protection systems andprotecting the integrity of copyright management informationsystems, including terms and conditions for use of a work.52 Theanti-circumvention provision of the DMCA prohibits (1) thecircumvention of technological measures that effectively controlaccess to or a work protected under U.S. copyright law and (2) themanufacture of, importation of, offering to the public of, providingof, or trafficking in any technology, product, service, device,component, or part thereof, that is primarily designed or producedfor the purpose of circumventing either anti-access or anti-copyingtechnological measures.53 The copyright managementinformation provision prohibits the unauthorized removal oralteration of copyright management information or providing,distributing, or importing for distribution, false copyrightmanagement information, which is defined to include any termsand conditions for use of the work.54The DMCA explicitly limits its prohibitions to circumventionof technological measures that effectively control access to or48. Id. § 2701(a). An “electronic communications service” is defined as “anyservice which provides to the users thereof the ability to send or receive wire orelectronic communications,” id. § 2510(15), while the term “electronic storage” isdefined as “any temporary, intermediate storage of a wire or electronic communicationincidental to the electronic transfer thereof . . . .” Id. § 2510(17).49. 15 U.S.C. § 1125(a) (1994).50. 17 U.S.C. § 1201 (Supp. IV 1998).51. The World Intellectual Property Organization Copyright Treaty, Apr. 12,1997, S. Treaty Doc. No. 105-17 (1997).52. 17 U.S.C. §§ 1201, 1202 (Supp. IV 1998).53. Id. § 1201.54. Id. § 1202.

copying of “a work protected under this title” [i.e., Title 17 of theU.S. Code], and removal or alteration of “copyright” managementinformation without the authority of the copyright owner.55 Thus,the DMCA would not seem to protect uncopyrightable databases.However, a party engaging in either circumvention oftechnological protection measures or deletion or alteration ofmanagement information associated with an ostensiblyuncopyrightable compilation of data apparently bears the riskthat the database will turn out to contain copyrightable subjectmatter, thereby subjecting the party to liability under the DMCA.Thus, the DMCA appears to provide inadvertent legal protectionfor self-help technological and contractual measures designed todeter data piracy.In addition to and complementing the DMCA, another potentlegislative source of protection for databases can be found in therecently promulgated Uniform Computer InformationTransactions Act (UCITA), which is explicitly designed to makeenforceable so-called “shrinkwrap” and “click-wrap” licenses.56UCITA has now been adopted in two states,57 thus creating theprospect that database contents can be effectively protected by acombination of technological protection measures and “click-wrap”contracts—in short, a kind of “electronic trade secret” protection.Even before UCITA was promulgated, the potency of this formof database protection had been illustrated in cases such asProCD, Inc. v. Zeidenberg,58 the facts of which are stronglyreminiscent of Feist. In ProCD, the plaintiff compiled more than3,000 telephone directories into a computer database and sold aversion of the database on CD-ROMs, accompanied by ashrinkwrap license encoded on the CD-ROMs, as well as printed55. Id. §§ 1201, 1202.56. Uniform Computer Information Transaction Act (UCITA), available athttp://www.law.upenn.edu/bll/ulc/ucita/ucitsFinal00.htm; see also, Amelia H. Boss,Taking UCITA on the Road: What Lessons Have We Learned?, 7 Roger Williams Univ.L. Rev. 179, 180 n.1 (2001).57. See Md. Code Ann., Com. Law II § 22 (Supp. 2001); Va. Code Ann. § 59.1-43(Michie 2001). For detailed information about the ongoing debate over the UniformComputer Information Transactions Act, see www.ucitaonline.com.58. 86 F.3d 1447 (7th Cir. 1996); see also Hill v. Gateway 2000, Inc., 105 F.3d1147 (7th Cir. 1997), cert. denied, 118 S. Ct. 47 (1997); Brower v. Gateway 2000, Inc.,676 N.Y.S.2d 569, 570 (N.Y. Sup. Ct. 1998); M.A. Mortenson Co. v. Timberline SoftwareCorp., 970 P.2d 803 (Wash. Ct. App. 1999); but see Step-Saver Data Sys., Inc. v. WyseTech., 939 F.2d 91 (3d Cir. 1991); Klocek v. Gateway, Inc., 104 F. Supp. 2d 1332 (D.Kan. 2000); U.S. Surgical Corp. v. Orris, Inc., 5 F. Supp. 2d 1201 (D. Kan. 1998).

in an instruction manual, which limited the use of the database tonon-commercial purposes.59 The defendant, Matthew Zeidenberg,bought a copy of the CD-ROM but decided to ignore the licenseand began reselling the information on the Internet to anyonewilling to pay a price that was less than what ProCD charged itscommercial customers for the database.60 ProCD filed suit,seeking an injunction against any further dissemination thatexceeded the rights specified in the license.61 The district courtrefused to grant relief, holding that the licenses were ineffectualbecause their terms did not appear on the outside of the CD-ROMpackages, and that even if the license terms were enforceablecontracts, they were preempted under section 301(a) of the UnitedStates Copyright Act.62 In an opinion by Judge FrankEasterbrook, the Court of Appeals for the Seventh Circuitreversed and remanded with instructions to enter judgment forthe plaintiff. The court of appeals held that the contracts wereenforceable under Article 2 of the Uniform Commercial Code andnot preempted by federal copyright law.63 Though the case iscontroversial, it nevertheless underscores the potency of existinglegal protections for databases under U.S. law.Meanwhile, the European Union, recognizing its shrinkingmarket share in the global information industry, promulgated itslandmark 1996 Database Directive requiring member countries toimplement sui generis protection that accords a broad,transferable right to prevent the extraction and/or utilization ofthe whole, or a substantial part (evaluated qualitatively and/orquantitatively), of the contents of a database, whether or notcopyrightable.64 The Directive makes it equally clear that themaker of a database may not prevent a lawful user fromextracting and/or re-utilizing insubstantial parts of the database,evaluated qualitatively and/or quantitatively, for any purposewhatsoever, and that any contractual provision to the contrary59. See ProCD, 86 F.3d at 1449-5060. Id. at 1450.61. Id.62. Id. at 1450, 1453.63. Id. at 1452-55. For a critique of the ProCD decision, see Charles R. McManis,The Privatization (or “Shrink-wrapping”) of American Copyright Law, 87 Cal. L. Rev.173, 178-79, 182-84 (1999).64. EU Database Directive, supra note 4, at art. 7.

shall be null and void.65 Member states may (but are not requiredto) create exceptions, allowing lawful users of a database madeavailable to the public to extract or re-utilize a substantial part ofits contents for three limited purposes.66 The term of protection isfifteen years from the first day of the next calendar year followingcompletion of the database, or if it is made available to the public,fifteen years from the first day of the next calendar year followingpublication.67 Any substantial change in the contents of thedatabase that would result in the database being considered asubstantially new investment would be entitled to its own term ofprotection.68 Finally, in stating that protection will be accorded todatabases whose makers are not nationals, habitual residents, orlegal persons established in a member state only if the country oforigin of the database offers “comparable protection” to databasesproduced by nationals, habitual residents or legal personsestablished in a member country of the European Community, theDirective seems to place those segments of the U.S. databaseindustry that do not have an established presence in the EU at apotentially serious competitive disadvantage in the EU.69Not surprisingly, promulgation of the EU Database Directiveimmediately generated demands for sui generis databaseprotection in the United States and stimulated a series ofproposed legislative responses. Thus far, however, the only resulthas been a legislative stalemate. As we have seen, the two billsmost recently considered by the United States Congress—H.R.354 and H.R. 1858—reflect two competing approaches to suigeneris database protection. H.R. 354 would accord a databasedeveloper a limited-term right,70 somewhat analogous to that65. Id. at art. 8.66. Id. at art. 9. Lawful users of a database made available to the public mayextract or re-utilize as substantial part of its contents: (a) in the case of extraction forprivate purposes of the contents of a non-electronic database; (b) in the case ofextraction for the purposes of illustration for teaching or scientific research, as long asthe source is indicated and to the extent justified by the non-commercial purpose to beachieved; and (c) in the case of extraction and/or re-utilization for the purposes ofpublic security or an administrative or judicial purpose. Id.67. Id. at art. 10.68. Id.69. See id. at art 11(3) and recital 56; but see infra note 184 and accompanyingtext.70. H.R. 354 § 1409(c), 106th Cong. (1999). For a detailed discussion of theinteresting way H.R. 354 manages to create a limited term of protection and yetmaintain its credentials as an antipiracy act see infra notes 168-71 and accompanying

created by the EU Database Directive, to prevent non-competingas well as competing uses of substantial portions of a database,thus recreating a limited form of the “sweat of the brow”protection laid to rest in Feist.71 By contrast, H.R. 1858 wouldprovide a narrower form of misappropriation protection byprohibiting only (1) the sale or distribution to the public of acompeting duplicate of a database collected and organized byanother and (2) misappropriation of real-time market information,thus effectively “re-federalizing” the holding in InternationalNews, at least as applied to databases.Other countries are closely following the debate in the UnitedStates over database protection. Japan may have opted in favor ofan unfair competition approach to database protection when itamended its Unfair Competition Law in 1993 to prohibit, for alimited three-year term of protection, the slavish imitation(referred to in Japanese as “dead copies”) of another’s goods.72While the subject of slavish copying under Japan’s UnfairCompetition Law can only be products, not services, computersoftware and databases apparently qualify as products whenmarketed on floppy discs or CDs, and the law may even extend tovirtual goods.73 Australia is said to be “walking a fine linebetween, on the one hand, supporting improvement of copyrightprotection against new uses in the digital environment and, on theother, avoiding excessive copyright protection resulting in denialof reasonable access to works and/or increased adverse balance oftrade in copyright royalties.”74 Korea is also considering howdatabases ought to be protected.75 Accordingly, the choice of oneU.S. bill over the other will be extremely influential intext.71. H.R. 354, § 1409(d).72. See Unfair Competition Act [Japan], Law No. 47/1993, § 2(3) (May 19, 1993);see generally Christopher Heath, The System of Unfair Competition Prevention inJapan 120-40 (2001) (discussing Japan’s approach to combating unfair competition).73. See Heath, supra note 72, at 130-31.74. Matilda in Cyberspace, Berne Protocol–Latest Development (July 9, 1996) (emailfrom Jaime Wodetski, commenting on database protection proposals prepared forthe benefit of the Libraries Copyright Committee), athttp://www.anu.edu.au/Matilda/issues/199607/715.source.html (last visited Apr. 11,2001).75. See Sang Jo Jong & Junu Park, Property versus Misappropriation: LegalProtection for Databases in Korea, 8 Wash. U. J.L. & Pol’y (forthcoming) (2002).75. See Sang Jo Jong & Junu Park, Property versus Misappropriation: LegalProtection for Databases in Korea, 8 Wash. U. J.L. & Pol’y (forthcoming) (2002).

determining the fate of database protection on the global scale.II. THE CHANGING FACE OF THE DATABASE INDUSTRYA. Databases and Digital TechnologyDigital technology is radically changing the face of thedatabase industry. The number of files in electronic databaseshas increased from four billion in 1991 to eleven billion in 1997, astaggering 200+% increase.76 The number of searches of suchinformation grew comparably, roughly doubling during any fiveyearperiod during the past decade.77 Likewise, the source ofdatabase development has dramatically shifted. In the late 1970s,78% of all databases were produced by government, academic andother non-profit providers, while the commercial sector accountedfor the remaining 22%. By 1997, the percentages were exactlyreversed.78While the database industry has long been producingdatabases in the absence of sui generis protection, the digitalrevolution has both enhanced the utility of databases and exposedthe vulnerability of investments in their production. Althoughprinted databases, such as telephone directories and retailcatalogs, remain important in certain industries, digitizedmultimedia databases have far more utility than printedcompilations. Because electronic databases can be navigated moreefficiently than printed publications, they are of growingimportance to the database industry and have made the creationof exhaustive factual compilations practical.79 Internet searchengines allow users to engage in their own selection andarrangement of the contents of mega-databases into smaller,individually tailored databases.The limitation of copyright to the original selection andarrangement of a database was developed with printed databasesin mind. The selection and arrangement of data in a printeddatabase creates value to the user because a random collection of76. O’Neal, supra note 2, at 109.77. Id.78. Id.79. See Kirk Statement, supra note 3, at 2 (noting that of the approximately11,000 databases listed in the Gale Directory of Databases, almost all are in electronicform—5,950 are made available online, 3,000 are on CD-ROM, 1,000 on diskette, and700 on magnetic tape).

facts would be of little or no use.80 Today, database softwareperforms the tedious task of coordination and arrangement.Internet databases are generally arranged to maximize storagecapacity and only become arranged for human use upon a user’sindividual request.Free riding in the database industry has been facilitated bythe exponential increase in the speed at which data can betransmitted. In 1991, the year Feist was decided, it would havetaken the average modem, working 24 hours a day, 154 days tocopy an average sized database.81 With current modems and theadvent of DSL, the same database can be copied in 1.7 minutes.82Electronic databases on the Internet afford many advantages,such as access and ease of subsequent modifications; however,accessibility also facilitates data piracy. A closed network wouldallow for a higher level of security and easier detection ofunauthorized uses, but is more difficult and time-consuming andmay tend to discourage use.83 As internet applications becomemore widespread, consumers expect access to data over the web.Moreover, the mere ability to protect a database with anticopyingtechnology does not necessarily obviate the need for legalprotection. The fact that a landowner can erect a fence around hisproperty does not mean he should forfeit a cause of action fortrespassing. The question currently before U.S. legislators is whatkind of fences should be created and what kind of trespass claimsshould be allowed.B. Reasons For Providing Sui Generis Protection of DatabasesThere are several reasons for providing sui generis protectionfor databases. First, the collection and organization of datarequire a large up-front investment. Second, the accessibility ofinformation in the digital information age has made the wholesalecopying of large compilations a simple task. Third, there isconsiderable concern among U.S. companies that the lack ofprotection in Europe, in the absence of reciprocal protection in theUnited States, will put them at a distinct competitivedisadvantage.80. Id. at 3.81. Id. at 5.82. Id.83. Id.

The compilation of a database requires a substantialinvestment. Producers of printed databases at one time enjoyednatural “lead-time” with which to exploit their product, as copyinga large printed compilation was time consuming. Today, thewholesale copying of an electronic database can be completed inseconds with 100% accuracy. The concept of providing artificial“lead-time” to minimize harm to a database developer’s market isreflected in the state common law “hot news” misappropriationdoctrine.84 Providing a limited term of such artificial leadtimeallows database developers to recoup their investment of researchand development costs.85 If free riders can immediately andperfectly copy the database, the original developer will be deprivedof the opportunity to develop its market niche.Because under current copyright law databases are protectedonly if and to the extent that they contain material selected orarranged in an organized manner, massive databases intended tobe comprehensive may fall short of the minimum level oforiginality required by Feist. Moreover, computer searchabledatabases are not organized in any humanly understandablemanner until there is a request from a specific user. Even if theorganization of a comprehensive database does present a modicumof creativity, the resulting “thin” copyright protection may beabrogated by the merger doctrine.86 Those works surviving themerger doctrine will be subject to a fair use privilege that isgenerally thought to be far broader when factual compilations,rather than highly creative works, are involved.87Economists argue that databases present a significantpotential for a free rider problem.88 For example, the American84. For a discussion of the origins of this doctrine, see supra notes 34-40 andaccompanying text.85. See generally J. H. Reichman & Pamela Samuelson, Intellectual PropertyRights in Data?, 50 Vand. L. Rev. 51, 145-46 (1997).86. See Morrissey v. Proctor & Gamble Co., 379 F.2d 675, 678-79 (1st Cir. 1967)(stating that when there is only a limited number of ways to express an idea, courtswill find that the idea has merged with the expression and thus not copyrightable).Highly factual or functional expression, such as recipes, parts lists, identificationnumbers, and the like, even if capable of being expressed in more than one way, areparticularly vulnerable to application of the merger doctrine.87. Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417 (1984).88. See, e.g., Collections of Information Antipiracy Act: Hearing on H.R. 2652Before the Subcomm. on Courts and Intellectual Prop. of the House Comm. on theJudiciary, 105th Cong. (1998) [hereinafter Hearing 2652] (statement of Robert E. Aber,on behalf of The NASDAQ Stock Market, Inc.), available at

Medical Association (AMA) licenses many of its databasesinternationally.89 If the EU reciprocity provision is enforced inthe absence of U.S. legislation, the AMA fears it may have littlerecourse for the widespread appropriation of its efforts inEurope.90 “Adopting database protection legislation in the UnitedStates could be a step toward providing the necessary ‘reciprocity’in order to protect U.S. databases in the EU.”91C. The Dangers of Too Much ProtectionAs compelling as these reasons may be for providing some sortof sui generis protection for databases, there are equal dangers inproviding too much protection. Recognizing the market effects ofintellectual property legislation is essential to the maintenance ofa sound industrial policy. The combination of sui generis databaseprotection, copyright law, unfair competition law, contract law andencryption measures may overprotect databases and thus allowdatabase developers to dominate markets and extract monopolyprofits.Opponents of sui generis protection for databases may wellargue that no further economic incentive is needed in the UnitedStates, as its database industry is not only thriving, butdominating the global market place under current U.S. law.Because copyright protection requires only a modicum ofcreativity in selection and arrangement, databases could probablybe organized sufficiently to qualify for at least “thin” copyrightprotection–though not for protection against wholesale duplicationof the data itself. As we have seen, database developers can alsorely on the contractual self-help measures of the sort authorizedby UCITA, which validates and enforces click-throughagreements.92 Also available to database developers are thetechnological self-help measures protected by the DMCA, whichprohibits both the circumvention of technological protection forcopyrighted works and the deletion or alteration of any copyrighthttp://www.house.gov/judiciary/41143.htm.89. Id. (statement of Richard F. Corlin, M.D., on behalf of the AMA), available athttp://www.house.gov/judiciary/41145.htm.90. For a less alarmist view of the availability of protection for U.S. databases inEurope, see infra note 184 and accompanying text.91. See Hearing 2652, supra note 88 (statement of Richard F. Corlin, M.D., onbehalf of the AMA), available at http://www.house.gov/judiciary/41145.htm.92. See supra notes 56-57 and accompanying text.

management information, including conditions for use of thework.93The scientific research community claims that sui generisdatabase legislation of the sort mandated by the EU DatabaseDirective and embodied in legislative predecessors of H.R. 354 willincrease costs for future research projects.94 Scientific researchersclaim that implementation of such legislation will make it moredifficult and costly to access data concerning a wide range of topicssuch as weather statistics and human genome data.95 Some datais so expensive that with sui generis database protection in effect,it would be cost-prohibitive to gather a large collection of facts.96One particularly striking example of the harm that can resultfrom a grant of property rights in data can be found in theprivatization of data from the Landsat series of remote sensingsatellites.97 Following privatization, the prices of Landsat dataincreased from approximately $400 to $4,400 per image,effectively putting them out of reach of most academics andindependent researchers.98More importantly, as one congressional witness noted: “To agreat degree, the value of technology is cumulative. We cannotmake progress without building freely on the data and results ofthe past.”99 Likewise, in testimony before the House ofRepresentatives Subcommittee on Courts and IntellectualProperty, a spokesman for several large scientific organizationsstated that:Data are the building blocks of knowledge and the seedsof discovery. They challenge us to develop new concepts,theories, and models to make sense of the patterns we see93. See supra notes 50-55 and accompanying text.94. See Andrew Lawler, Database Access Fight Heats Up, Science, Nov. 15, 1996,at 1074.95. Id.96. See, e.g., Anne Linn, History of Database Protection: Legal Issues of Concern tothe Scientific Community (Mar. 3, 2000) (“a single synthetic aperture radar scene costs$1600.”), at http://www.codata.org/codata/data_access/linn.html (last visited Apr. 11,2001).97. See Michael J. Bastian, Protection of “Noncreative” Databases: Harmonizationof United States, Foreign and International Law, 22 B.C. Int’l & Comp. L. Rev. 425, 431(1999).98. Id. at 431-32.99. Hearing 2652, supra note 88 (statement of Tim D. Casey on behalf of theInformation Technology Association of America), available athttp://www.house.gov/judiciary/41149.htm.

in them. They provide the quantitative basis for testingand confirming theories and for translating newdiscoveries into useful applications for the benefit ofsociety. They also are the foundation of sensible publicpolicy in our democracy. The assembled record ofscientific data and resulting information is both a historyof events in the natural world and a record of humanaccomplishment.100When assessing the value of sui generis legislation and deciding towhat degree exclusionary rights should be granted, lawmakersmust balance the social benefits of databases that would not beavailable in the absence of protection against the loss ofinnovative and transformative uses of databases that might notoccur because of the existence of database protection. Thescientific community and research and development industriesfrequently conduct research by applying novel theories of analysisto existing data in order to highlight certain trends. By reviewinglarger databases and carefully organizing selected data, aresearcher may discover trends lost in the larger context of theoriginal database. By further refining the original data, secondarydatabases thus add value to the original.101Some sectors of the information technology industry itself arealso disturbed by the prospect of sui generis protection fordatabases. They claim that the imposition of increased businesscosts will drive the entire information technology industryoffshore and heighten the dangers of incipient monopoly.102 As100. Collections of Information Antipiracy Act: Hearing on H.R. 354 before theSubcomm. on Courts and Intellectual Prop. of the House Comm. on the Judiciary, 106thCong. (1999) (statement of Joshua Lederberg, Nobel laureate, on behalf of NAS, NAE,IOM and the American Association for the Advancement of Science), available athttp://www.house.gov/judiciary/106-lede.htm.101. National Research Council, Preserving Scientific Data on Our PhysicalUniverse 16 (1995):In all areas of research, the collection of data sets is not an end in itself, butrather a means to an end, the first step in the creation of new information,knowledge, and understanding. As part of that process, the originaldatabases are continually refined and recombined to create new databasesand new insights. Typically, each level of processing adds value to an original(raw) data set by summarizing the original product, synthesizing a newproduct, or providing an interpretation of the original data.Id.102. Hearing 2652, supra note 88 (statement of Tim D. Casey on behalf of theInformation Technology Association Of America), available at

one congressional witness noted: “Tom [Feist] was left with nochoice but to copy listings in order to provide consumers aconvenient, one-book directory covering eleven different serviceareas, because one of the telcos refused to license its listings tohim.”103 Internet based companies argue that database protectionwill result in industry concentration, and will increase the costand diminish the utility of search engines, as web browsers willeither have to maintain their own databases or license them fromcompetitors.104 Technologically oriented companies also predictrising costs associated with world wide web industries due tomonopoly pricing.105 Concern has also been expressed that“granting protection for interface specifications could threatencompetition on the Internet” by concentrating a highly competitiveindustry into a nucleus of dominant firms.106Overly broad database protection could thus inhibit theproduction of transformative databases by raising barriers toentry and costs associated with gathering, verifying andmaintaining compilations.107 While observations and facts havelittle inherent value, the collection and processing of such facts todiscern trends adds economic value to the bare facts. Unprocessedfacts are the most difficult to understand and frequently of littleuse to anyone other than an expert researcher.108 “With everysuccessive level of processing, the data tend to become moreunderstandable and frequently are better documented for the layuser.”109 As a matter of sound public policy, scientists should beencouraged to analyze and refine, rather than simply collect, data;protection should only be accorded to compilations that add valueto the underlying facts by providing a new or creativeorganization. Regard for the value of transformative uses ofdatabases reveals an overlap with copyright protection. As rawhttp://www.house.gov/judiciary/41149.htm.103. Id. (statement of William Hammack on behalf of the Association of DirectoryPublishers), available at http://www.house.gov/judiciary/4116.htm.104. See generally id. (discussing the impact of concentrated ownership) (statementof Jonathan Band on behalf of the Online Banking Association), available athttp://www.house.gov/judiciary/41148.htm.105. Id.106. Id.107. See Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 579 (1994).108. See J. H. Reichman & Paul F. Uhlir, Database Protection at the Crossroads:Recent Developments and Their Impact on Science on Technology, 14 Berkeley Tech.L.J. 793 (1999).109. Id. at 813.

facts are selected, refined and arranged, their expression is morelikely to benefit from copyright protection.Because sui generis protection will create an economic barrierto the access of information, not only companies but alsodeveloping countries would inevitably face higher barriers toentry. The worldwide implementation of protective legislation fordatabases could have profound economic consequences fordeveloping countries. Because the right to exclude is inherent inintellectual property rights, “enhancing the value of theintellectual property assets of one company—or country—willnecessarily lead to increased intellectual property liability of othercompanies or countries.”110D. Term of ProtectionA further problem posed by sui generis proposals for databaseprotection is defining the term of protection. There are threepossible schemes for defining the term of protection for adatabase. The first provides static protection for a term beginningat the completion of the original edition of the database. Thesecond provides dynamic protection for the most recent edition ofthe database in its entirety. The third provides variable protectionfor the original database and subsequent additions for staggeredfixed terms.The static model protects an initial investment, but does notprovide incentives for the developer to update the compilation.The static model has an inherent bias in favor of historic data, asupdates to databases would not be protected. Dynamic protectionencourages further revision and development of a database bygranting a new term of protection for each new edition of thework. A major concern regarding dynamic protection forelectronic databases is that the term may effectively be perpetualfor frequently updated databases, unless some method ofdistinguishing between new and old material is specified.111 Thevariable model provides a fixed term of protection to the originalwork and protects new entries in future editions for a similarterm, resulting in staggered terms of protection for differententries in the same database. Protecting only those modifications110. Charles R. McManis, Intellectual Property and International Mergers andAcquisitions, 66 U. Cin. L. Rev. 1283, 1289 (1998).111. See infra note 132 and accompanying text.

that were made after the initial term expires may be impractical.Either the database developer would have to provide notice, whichwould create unnecessary bulk and complexity within thedatabase, or the user would incur the burden of identifying whichentries were protected and which were not.The very difficulty in developing a workable option regardingthe term of a property right in data suggests that a liabilityscheme may be preferable.III. THE EUROPEAN DIRECTIVEA. Impetus for the EU DirectiveThe dilemma for Europe is that the United States currentlydominates the world database industry.112 While in 1990European database developers produced almost half the worldtotal of databases, they accounted for only one-quarter of the totalrevenue.113 The explanation for the discrepancy is that non-profitproducers (i.e., governments) produced the majority (54%) ofEuropean databases, whereas the private sector produced themajority of databases in the United States.114 With one exception,Europe is nearly completely dependent on foreign commercialdatabases.115 The exception is Reuters, a U.K. corporation, whichis the world’s largest supplier of real-time financialinformation.116 Given this state of affairs, it is initially surprisingthat the EU would be the first to implement highly protectionistdatabase legislation. There are however basically two underlyingreasons—one technological and one linguistic.1. Development of Computer Related TechnologyAs the market for printed databases deteriorates, the112. Mortimer B. Zuckerman, The Times of Our Lives, U.S. News & World Rep.,Dec. 27, 1999, at 68 (stating that of the “48 information technology companies thatMorgan Stanley Dean Witter believes will enjoy a competitive advantage over the nextnumber of years, 31 are American. Only six are European.”).113. See Charles von Simson, Note, Feist or Famine-American Database Copyrightas an Economic Model for the European Union, 20 Brook. J. Int’l L. 729, 731 (1995).114. Id.115. See Information Market Observatory, European Information Trends, § 1, athttp://158.169.50.95:10080/imo/en/trend96/trend96-toc.html (last visited Dec. 11, 2001).Iceland imports 93.3% of data, Austria imports 86.2%, and Luxembourg imports 88.0%.Id.116. Id. § 3.5.

integration of modern computing technologies becomesincreasingly important.117 The European Union sought legislativeprotection for databases in an effort to reverse a growing tradedeficit in the electronic data market. The promulgation of theDatabase Directive came on the heels of a 1994 study of electronicinformation services, which stated that, “with the exceptions ofthe United Kingdom and the Netherlands all EEA countries hadnegative trade balances.”118The digital revolution originated in the United States ratherthan in Europe, and the European Union is thus getting a latestart. The Internet was created by the United States governmentand quickly fostered “a thriving Internet core economy, creatingnew businesses, new revenue streams, and, more importantly,new jobs.”119 As of 1997, the United States was acknowledged tohave built a substantial lead over Europe.120 The EU alsoacknowledged that a similar lead seemed to be opening up in thestrategic sector of electronic commerce tools, products andtechnologies that underpin the future development of electroniccommerce.1212. Linguistic FragmentationLinguistic fragmentation may be another reason that theEuropean market share of the database industry is eroding. TheEU operates in eleven different languages; every documentproduced by EU government authorities is translated into eachlanguage before publication.122 The large number of languagescomplicates competitive markets by eliminating economies of scalein the production of language-based databases. As English is the117. Of 147 European directory publishers, only 27 earned gross revenues of 50million ECU per year. These publishers are also responsible for many retail catalogs.As corporate catalogs are indicative of future retail performances, the potential loss ofcatalog publication companies within the EU may make it easier for foreign firms todevelop a significant market share. Id.118. Id. § 2.2.119. A European Initiative in Electronic Commerce, Communication to theEuropean Parliament, the Council, the Economic and Social Committee and theCommittee of the Regions, at para. 17, available athttp://www.cordis.lu/src/ecomcom1.htm (last updated April 16, 1997).120. Id. at para. 18.121. Id.122. See, e.g., http://europa.ei.int. The official European Union website is availablein 11 different languages.

preferred language of many business transactions, the demand ishigher for English-based databases.Between 1992 and 1998 the percentage of sites based onEnglish rose from 67% to 70%.123 The worldwide number ofinternet sites based on the English language is currently 65.1%.124As databases are frequently based in part on prior datacompilations, the entrenchment of English language databasesshould continue. Additionally, the two largest encyclopediaservices, Microsoft’s Encarta and the British EncyclopediaBritannica, are English-based works.Nearly half of the current 250 million Americans rely on internetbased data.125 That consideration alone accounts for the highdemand for data provided in the English language. Moreover,recent economic growth in the United States has encouragedfurther capital investment in the information technologyindustry.126 The United States is also home to many of theworld’s major software producers, including AT&T, HewlettPackard, Microsoft, Motorola, IBM, GTE, Lucent Technologies andOracle. As software is important to the creation of databases, theplatform on which they are created is most often based on EnglishB. The EU Database Directive and Its ImplementationThe EU Database Directive, promulgated on March 11, 1996,mandated implementation by all states in the European EconomicArea by January 1, 1998.127 As we have seen, the EU DatabaseDirective both defines the scope of copyright protection fordatabases and creates a new sui generis right. Consistent withArticle 10 of TRIPS, the Directive specifies that copyrightprotection extends to databases that by reason of the selection orarrangement of their contents, constitute the author’s ownintellectual creation.128 The exclusive rights granted to qualifyingdatabases largely parallel the exclusive rights granted under U.S.123. Equally significant is the numbers that underlie those percentages. In 1992,1,489 of 2,213 multimedia databases were in English. By 1998 the number ofmultimedia databases had risen to 28,199, of which 19,602 were in English. Id. at 31.124. Id. at 4.125. Id.126. Information Market Observatory, supra note 115, § 4.1.127. EU Database Directive, supra note 4, at art. 16, § 1.128. Id. at art. 3.

copyright law.129 The sui generis rights are vested in “a maker ofa database which shows that there has been qualitatively and/orquantitatively a substantial investment” in its creation, and allowthe maker to prevent the extraction and/or re-utilization of thecontents in whole or in any substantial part (evaluatedqualitatively and/or quantitatively).130 Although the initial termof the sui generis right is fifteen years from the date of completionof a database, the EU Directive mandates a new term of protectionfor any database in which the original contents undergosubstantial change, and imposes no obligation to identify whatdata is new and what has ostensibly fallen into the publicdomain.131 The effect of this provision is to create a perpetuallyexclusive right for dynamic databases.132 Also troubling is thatthe EU Directive authorizes but does not require member states tocreate fair use exceptions to the general prohibition.133Two reasons have been given to explain why the EU adopteda sui generis property right rather than a norm of unfaircompetition—(1) the logistical difficulty of harmonizing unfaircompetition laws across the board and (2) a desire to protectdatabases from “information Samaritans” as well as free riders.134Some commentators have observed, however, that the creation of129. Id. at art. 5. The Directive provides:The author of a database shall have the exclusive right to carry out or authorize:(1) temporary or permanent reproduction by any means and in any form, inwhole or in part(2) translation, adaptation, arrangement and any other alteration(3) any form of distribution to the public of the database or copies thereof(4) any communication, display or performance to the public(5) any reproduction, distribution, communication, display or performance tothe public of the results of the acts referred to in (b).Id.130. Id. at art. 7, § 1.131. Id. at art. 10, §§ 1, 2.132. The danger of inadvertently creating perpetual protection for databases in theUnited States was duly noted. See The Collections of Information Antipiracy Act:Hearings on H.R. 354 Before the Subcomm. on Courts and Intellectual Prop. of theComm. on the Judiciary, 106th Cong. (Mar. 19, 1999) (statement of Andrew J. Pincus,General Counsel, United States Department of Commerce) [hereinafter PincusStatement], available at http://www.house.gov/judiciary/106-pinc.htm.133. EU Database Directive, supra note 4, at art. 9; see also supra note 66 andaccompanying text.134. See Michael J. Bastian, supra note 97, at 444; Mark Powell, The EuropeanUnion’s Database Directive: An International Antidote of the Side Effects of Feist?, 20Fordham Int’l L.J. 1215, 1224-25 (1997); Reichman & Samuelson, supra note 85, at 81.

the sui generis right mainly serves its intended purpose to “favorEuropean database publishers at the expense of their customersand non-EU competitors.”135 Others have noted a discrepancybetween the economic arguments justifying the protection ofdatabase contents and the particular solution proposed in theDirective.136 The main reason given for a “copyright plus”approach to database protection is to prevent the potentialdamage caused by the slavish copying by competitors; the“economic case for the creation of a right to prevent extraction andreutilization of unoriginal content by users has never beensatisfactorily explained.”137 Even so, the compromise reached inthe Database Directive has satisfied neither database makers norusers.138 Whereas users would have preferred to see compulsorylicensing provisions incorporated into the Directive, rather thansimply held in reserve by the review clause of Article 16, databasemakers would have liked, among other things, to prevent theextraction and re-utilization of insubstantial amounts of thecontents of databases, if not by law, then by contract.139As of the end of 1998, only nine member states had passednational legislation.140 The Commission of the European Unionwas forced to file complaints in the European Court of Justiceagainst several member states.141 Today, with the exception ofIreland,142 all member states are now in compliance with theDirective.143 However, the Directive has yielded disappointingresults.144 Harmonization of the originality standard for135. See von Simson, supra note 113, at 735.136. Powell, supra note 134, at 1225.137. Id.138. Id. at 1217.139. Id.140. Linn, supra note 96, at Table 1. The nine states were: Austria, Jan. 1, 1998;Belgium, Sept. 1, 1998; Denmark, July 1, 1998; Finland, Apr. 4, 1998; France, June 16,1998; Germany, Jan. 1, 1998; Spain, Apr. 1, 1998; Sweden, Jan. 1, 1998; UnitedKingdom, Jan. 1, 1998.141. Countries with no implementation legislation as of January 1, 1999: Greece,Ireland, Italy, Luxembourg, The Netherlands, Portugal, Iceland, Norway andLiechtenstein. See, e.g., Case C-484/99, Comm’n v. Hellenic Republic (1999) 2000 O.J.(C47/24); Case C-506/99, Comm’n v. Portuguese Republic (1999) 2000 O.J. (C79/13).142. See ECJ C-370/199, Comm’n v. Ireland, 11 January 2001.143. See F.W. Grosheide, Database Protection-The European Way, 8 Wash. U. J.L.& Pol’y (forthcoming) (2002).144. P. Bernt Hugenholtz, Implementing the European Database Directive, inIntellectual Property and Information Law: Essays in Honour of Herman CohenJehoram 194 (Jan C. Kabel & Gerard J.H.M. Mom eds., 1998).

copyright protection has not been achieved. Moreover, in somecountries compilations of data are now subjected to a triplestandard of protection, as countries which previously granted ameasure of protection to mere compilations of data have invokedtheir right under Recital 52 of the Directive to retain exceptionstraditionally specified by such rules.145 Copyright limitationsapplicable to databases continue to vary from country to country,as do, albeit to a lesser degree, the limitations on the sui generisright.C. ReciprocityThe European Database Directive contains a reciprocityprovision that only accords protection to non-resident foreignerswhose home countries have enacted similar legislation.146 Whilethis provision discriminates on the basis on nationality, theDirective was carefully drafted to avoid conflict with the national(i.e., non-discriminatory) treatment provisions of both TRIPS andthe Berne Convention.147 As the United States is now learning,allowing countries with sufficient market power to force suigeneris intellectual property provisions on other countries throughreciprocity provisions creates a dangerous precedent. As recentexperience illustrates, industrialized nations can use reciprocityprovisions to influence the public policy choices of other countries.Further, developing countries lacking sufficient market power toimpose legislation through reciprocity provisions will likely befurther disadvantaged relative to industrialized nations.It has been argued elsewhere that the new, sui generis dataright (like the semiconductor chip design right that preceded it)might fall within the meaning of “industrial property” as used in145. EU Database Directive, supra note 4, at Recital 52.146. Id. at Recital 56, which states:[T]he right to prevent unauthorized extraction and/or re-utilization of adatabase should apply to databases whose makers are nationals or habitualresidents of third countries or to those produced by legal persons notestablished in a Member State, within the meaning of the Treaty, only if suchthird countries offer comparable protection to databases produced bynationals of Member States or persons who have their habitual residence inthe territory of the Community.Id.147. Debra B. Rosler, The European Union’s Proposed Directive for the LegalProtection of Databases: A New Threat to the Free Flow of Information, 10 High Tech.L.J. 105, 137 (1995).

the Paris Convention for the Protection of Industrial Property;thus, even if data rights fall outside the national treatmentprovisions of TRIPS and the Berne Convention, the sui generisdata right may be subject to the national treatment provisions ofthe Paris Convention.148 However, until this question is raisedand resolved, either in World Trade Organization (WTO) disputesettlement proceedings or elsewhere, the United States mustmake a unilateral determination as to what is in its nationalinterest.IV. UNITED STATES LEGISLATIVE EFFORTSA. History of Database Protection in the United StatesWithin ten weeks of the promulgation of the EU DatabaseDirective, the first legislation aimed at creating a sui generis rightin the United States, H.R. 3531, was proposed in the House ofRepresentatives.149 H.R. 3531 would have created a very strongproperty right with a twenty-five year term and potentially severecriminal sanctions for infringement.150 Opposition wasunexpectedly intense and the bill never reached the floor. H.R.3531’s sequel, H.R. 2652, responded to concerns from the researchand library communities by including limited fair useprovisions.151 This bill was passed by the House as part of theDigital Millennium Copyright Act on August 4, 1998; however, itwas subsequently stricken during negotiations with the Senate,and the DMCA passed shortly thereafter on October 28, 1998.152The Collections of Information Antipiracy Act, H.R. 354, asintroduced on January 19, 1999, would accord a databasedeveloper a broad right in factual compilations, subject not only toa specific limitation for certain non-profit educational, scientificand research uses, but also to a broader limitation for “reasonableuses” generally.153 The reasonableness of a particular use is to be148. See McManis, Taking TRIPS, supra note 16, at 258-59; Paul E. Geller,Intellectual Property in the Global Marketplace: Impact of TRIPS Dispute Settlements?,29 Int’l. Law. 99, 112 (1995).149. Database Investment and Intellectual Property Antipiracy Act of 1996, H.R.3531 §§ 6, 9, 104th Cong. (1996).150. See id.151. Collections of Information Antipiracy Act, H.R. 2652, 105th Cong. (1998).152. Digital Millenium Copyright Act, 17 U.S.C. § 1201 (Supp. IV 1998).153. H.R. 354 § 1403, 106th Cong. (1999).

evaluated by a multi-factor test that is somewhat reminiscent ofthe fair use privilege contained in section 107 of the United StatesCopyright Act of 1976.154 Notwithstanding the inclusion of thisbroadened fair use provision, opponents of H.R. 354 introduced analternative bill on May 19, 1999, the Consumer and InvestorAccess to Information Act, H.R. 1858, that based liability on atheory more analogous to the common law tort of “hot news”misappropriation.155B. Comparison of H.R. 354, H.R. 1858 and the EU DatabaseDirective1. Theories and Scope of ProtectionAs we have seen, H.R. 354 creates a sui generis right in thecontents of a database, somewhat analogous to that mandated bythe EU Database Directive, thus recreating a limited form of the“sweat of the brow” protection laid to rest in Feist.156 By contrast,H.R. 1858 protects databases by creating liability formisappropriation, thereby “re-federalizing” the United StatesSupreme Court’s holding in International News, at least asapplied to databases.157H.R. 354 prohibits the extraction or making available toothers all or a substantial part of a collection of informationgathered, organized or maintained by another person through theinvestment of substantial resources, that causes material harm tothe primary or a related market of that other person or of asuccessor in interest.158 The key to understanding the scope ofthis property right, and how it differs from a right based on atheory of unfair competition, is in the definitions of “primary” and“related” markets.A primary market is any market in which a product or servicethat incorporates a collection of information is offered and inwhich a person claiming protection with respect to that collectionof information derives or reasonably expects to derive revenue—154. 17 U.S.C. § 107 (1994).155. H.R. 1858 § 103, 106th Cong. (1999).156. See supra note 71 and accompanying text.157. See supra note 71 and accompanying text.158. H.R. 354 § 1402, 106th Cong. (1999).

directly or indirectly.159 In other words, a primary market is anexisting market in which the person claiming protection alreadyearns or reasonably expects to derive revenue. A related market,by contrast, includes each of the following (1) any market in whicha person claiming protection with respect to a collection ofinformation has “taken demonstrable steps” to offer in commercewithin a short period of time a product or service incorporatingthat collection of information with the reasonable expectation toderive revenue, directly or indirectly (i.e., a “zone of probableexpansion,” to borrow a phrase from trademark law) and (2) anymarket in which products or services that incorporate collectionsof information similar to a product or service offered by the personclaiming protection are offered and in which persons offering suchsimilar products or services derive or reasonably expect to deriverevenue, directly or indirectly (i.e., virtually any potential marketin which a compilation of information or substantial portionthereof might reasonably be expected to earn revenue).160The original language of the bill would have extended itsprotection to any “actual or potential market” of a database, anddefined a “potential market” as any market in which a claimantfor protection “has current and demonstrable plans to exploit orthat is commonly exploited by persons offering similar products orservices incorporating collections of information.”161 The ClintonAdministration was sufficiently concerned over the breadth of thisproposed language that in testimony before the HouseSubcommittee on Courts and Intellectual Property the GeneralCounsel of the United States Department of Commerce, Andrew J.Pincus, suggested that the Subcommittee reexamine the conceptsof “actual” and “potential” markets.162 Yet, apart fromsubstituting the word “related” for the word “potential,” the finalbill seems not to have responded to the Administration’s concerns.As we have seen, the EU Database Directive limits thepermissible exceptions to the sui generis right in electronicdatabases to extraction for the purposes of illustration forteaching or scientific research, and only so long as the source isindicated and to the extent justified by the non-commercial159. Id. §§ 1401(3)-(4), 1402.160. Id. § 1401(4).161. See H.R. 354 §§ 1401(3), 1402, 106th Cong. (Jan. 19, 1999) (as originallyintroduced in the House of Representatives).162. See Pincus Statement, supra note 132, at 5.

purpose to be achieved.163 By contrast, H.R. 354 not only containsa specific privilege for certain non-profit educational, scientificand research uses164 (as well as a variety of other specificprivileges165), but also contains a much broader “reasonable use”privilege,166 which is reminiscent of the fair use privilegecontained in section 107 of the United States Copyright Act.While this “reasonable use” provision was apparently considered anecessary concession to domestic critics of earlier versions of thebill, the resulting difference in the scope of permissible uses underH.R. 354 and the EU Database Directive raises the questionwhether H.R. 354 would in fact provide protection “comparable” tothe sui generis right mandated in the EU Directive.The question of comparability becomes still murkier when theterms of protection under H.R. 354 and the EU Database Directiveare compared. While both provide for a fifteen year term ofprotection that is subject to extension in order to protectcumulatively substantial updating of the database, the EUDatabase Directive apparently mandates a dynamic approach thatwould protect a new edition in its entirety, whereas H.R. 354adopts a variable term of protection that would only protect thenew material in the new edition.167Article 10 of the EU Database Directive states that anysubstantial change to the contents of a database that causes thedatabase to be considered a substantial new investment qualifies“the database resulting from that investment”—and not simplythe new material—to its own term of protection. The databaseproducer is apparently under no obligation to distinguish whatdata in that database is new and what has been previouslyprotected.By contrast, H.R. 354 states (1) that no action may “bemaintained for making available or extracting all or a substantialpart of a collection of information that occurs more than 15 yearsafter the portion of the collection that is made available orextracted was first offered in commerce following the investment163. EU Database Directive, supra note 4, at art. 9(b).164. H.R. 354 § 1403(b), 106th Cong. (1999).165. Id. § 1403(c)(i) (allowing individual items and other insubstantial parts,independently gathered information, verification purposes, news reporting, transfer ofcopy, genealogical information, and investigative, protective, or intelligence activities).166. Id. § 1403(a).167. Id. § 1409(c).

of resources that qualifies that portion of the collection forprotection,”168 and (2) that “the person claiming protection bearsthe burden of proving “that the date on which the portion of thecollection that was made available or extracted was firstoffered . . . in commerce . . . was no more than fifteen years priorto the time when it was made available or extracted by thedefendant.”169 H.R. 354 also states that “[n]o monetary reliefshall be available . . . [if the defendant] could not reasonablydetermine whether the date on which the portion of the collectionthat was made available or extracted was first offered incommerce” was more than 15 years prior to the violation.170Where “a collection of information into which all or a substantialpart of a government collection of information isincorporated, . . . no monetary relief shall be available for aviolation . . . unless a statement appear[s] . . . in a manner andlocation so as to give reasonable notice, identifying thegovernment collection and the government entity from which itwas obtained.”171In short, whereas the EU Database Directive appears tomandate dynamic (i.e., continuing) protection for an entiredatabase that is the subject of cumulatively substantial updating,H.R. 354 would protect only those substantial portions of adatabase that can be shown to have been offered in commercewithin the last fifteen years, and would deny monetary reliefwhere the defendant could not reasonably distinguish old and newdata, or in the case of collections incorporating informationcollected by the government, was not given reasonable notificationof the government source. Here, too, the difference in the scope ofprotection mandated by the EU Directive and that provided forunder H.R. 354 raises the question whether the European Councilwould consider the two to be comparable.2. H.R. 1858In contrast to both H.R. 354 and the EU Database Directive,H.R. 1858 is a much more narrowly tailored bill, which would (1)prevent wholesale competitive duplication of databases and (2)168. Id.169. Id. § 1409(d).170. H.R. 354 § 1408(a), 106th Cong. (1999).171. Id. § 1408(b).

prohibit misappropriation of real-time market information.172Title I of H.R. 1858 would make it a violation to sell or distributeto the public a database that is a duplicate of another database,when the duplicate is sold or distributed in competition with theoriginal database.173 The Federal Trade Commission (FTC) isvested with specific jurisdiction, under section 5 of the FederalTrade Commission Act,174 to prevent violations of the proposedact,175 but the bill does not expressly create a private right ofaction to enforce violations of section 102.This limited enforcement mechanism is H.R. 1858’s mostcontroversial feature. While the Federal Trade Commissionclaims to have experience in formulating policy and remedies inthis field176 and has agreed that the threat of private actions couldbe used by market incumbents to threaten potential entrants,thus potentially raising difficult issues for courts called on tointerpret the misuse defense contained in section 106(b) of thebill,177 the FTC itself has expressed concern about theenforcement feature of the bill, particularly if the FTC is the solestatutory enforcer.178 Apparently, the drafters of H.R. 1858 do not172. H.R. 1858 §§ 102, 201, 106th Cong. (1999).173. Id. § 102(1), (2).174. 15 U.S.C. § 45 (1997).175. Id. § 107(c). A recent FTC consent agreement and order suggests that the FTCmay already have jurisdiction under section 5 to prevent this form of unfaircompetition. See Stipulated Consent Agreement and Final Order, FTC v.ReverseAuction.com, Inc. (DDC 2000) (stating that under the threat of FTC action,ReverseAuction entered into a consent decree for violating eBay’s “user agreement,”because ReverseAuction, an upstart competitor in the online auction market, copiedeBay’s customer database in order to solicit new customers), athttp://www.ftc.gov/os/2000/01/reverseconsent.htm (last visited Dec. 27, 2001).176. See Consumer and Investor Access to Information Act of 1999: Hearing on H.R.1858 Before the Subcomm. on Telecommunications, Trade and Consumer Protection,Comm. on Commerce, 106th Cong. (1999) (prepared statement of the Federal TradeCommission) [hereinafter FTC Prepared Statement], available athttp://www.ftc.gov/opp/hr09881.htm. In their prepared statement the FTC cites theHearings on Global and Innovation-Based Competition, and three of its recentdecisions, In re Softsearch Holdings, Inc., 5 Trade Reg. Rep. (CCH) 24,171 (F.T.C.July 28, 1997 (consent decree); In re Automatic Data Processing, Inc., 5 Trade Reg.Rep. (CCH) 24,006 (F.T.C. Mar. 27, 1996) (consent decree); In re Provident Cos., Inc.,No. 991-0101, 64 Fed. Reg. 27,991 (F.T.C. May 24, 1999) (proposed consent decree,subject to public comment), in which the FTC considered the potentiallyanticompetitive effects of the increased market power that can result fromconsolidation among database owners and vendors. Id.177. FTC Prepared Statement, supra note 176, at 10; see also infra notes 188-89and accompanying text.178. Id. In a footnote, the FTC states its understanding that state common law

wish to encourage private lawsuits, and believe that marketincumbents, particularly sole-source providers, will think twicebefore complaining about data piracy to one of the country’s twofederal antitrust enforcement agencies.Although Title I would be enforceable only through the FTCproceedings, the theory of liability is akin to the common law tortof misappropriation of “hot news.”179 Value-added andtransformative uses are permitted, as H.R. 1858 considers a copyto be an infringement only if it is “substantially the same” as theoriginal,180 and prohibits the public sale or distribution of aduplicate database only where it is distributed in competition withthe original database.181Title II of H.R. 1858 would create a separate prohibitionagainst the “misappropriation” of real-time market information.Section 201 states that any person who “obtains directly orindirectly from a market information processor real-time marketinformation,” and “directly or indirectly sells, distributes,redistributes, or otherwise disseminates such information withoutthe authorization of such market information processor, shall beliable to that market information processor . . . .”182 In contrast toTitle I, the prohibition of Title II is not limited to the wholesaleduplications of real-time market information, nor to the sale ordistribution of the information in competition with theinformation processor. Moreover, this prohibition would beprivately enforceable through a civil action for monetary andinjunctive relief.183While the protection that H.R. 1858 would provide couldhardly be called comparable to that specified in H.R. 354 ormandated by the EU Database Directive, this is not to suggestthat enactment of H.R. 1858, rather than H.R. 354, would leaveU.S. database producers wholly unprotected in Europe. To themisappropriation suits involving databases will generally be preempted under section105(b) of the proposed act. Id. at 10 n.68. However, section 105(b) merely states that“no State law that prohibits or that otherwise regulates conduct that is subject to theprohibitions specified in section 102 shall be effective to the extent that such State lawis inconsistent with section 102.” H.R. 1858 § 105(b), 106th Cong. (1999) (emphasisadded).179. See supra notes 34-45 and accompanying text.180. H.R. 1858 §§ 101(2), 102.181. Id. § 102.182. Id. § 201 (proposing to amend § 11A of the Securities Exchange Act of 1904).183. Id.

contrary, U.S. database producers would arguably be entitled toprecisely the same protection in Europe that they would be (and infact currently are) entitled to in the United States. This is sobecause Article 10bis of the Paris Convention states that membercountries of the Paris Union are bound to assure nationals of suchcountries effective protection against unfair competition, and goeson to state that any act of competition contrary to honest practicesin industrial or commercial matters constitutes an act of unfaircompetition.184 Given that the EU has now mandated sui generisprotection for database contents even against non-competitiveextractions and re-utilizations, European countries can hardlyargue that wholesale duplication of database contents in order tocompete with the database producer is not an act of unfaircompetition within the meaning of Article 10bis of the ParisConvention.Thus, the argument that enactment of sui generis protectionof the sort spelled out in H.R. 354 is necessary (or sufficient) forU.S. database producers to obtain protection for their databases inthe EU is suspect. As we have also seen, it is not clear thatpassage of H.R. 354 will be deemed sufficiently comparable toqualify U.S. database producers for the (arguably over-broad) suigeneris protection now available in Europe. Indeed, one featurethat H.R. 354 and H.R. 1858 have in common distinguishes bothbills from the type of protection mandated by the EU Directive,and further undermines the argument that either bill wouldprovide comparable protection. This feature concerns theenforceability of contracts that put additional restrictions on theuse of database contents.3. Contractual CircumventionWhile the two U.S. bills provide two different approaches tothe protection of database contents, both bills make it clear thatnothing in the legislation would preempt contracts that placeadditional restrictions on the use of a database. H.R. 354 isparticularly adamant in this regard. Section 1405(a) of H.R. 354states that nothing in this bill “shall affect rights . . . orobligations relating to information, including . . . the law of184. Paris Convention for the Protection of Industrial Property, Mar. 20, 1883, 13U.S.T. 2, 828 U.N.T.S. 107, as last revised at the Stockholm Revision Conference, July14, 1967, 21 U.S.T. 1538, 828 U.N.T.S. 303, art. 10bis.

contract.”185 While section 1405(b) would preempt state law withrespect to rights that are equivalent to those protected under thenew act, that same section makes it clear that the law of contracts“shall not be deemed to provide equivalent rights for purposes ofthis subsection.”186 Additionally, section 1405(e) reiterates thatnothing in the proposed act “shall restrict the rights ofparties . . . to enter into licenses or any other contracts withrespect to making available or extracting collections ofinformation.”187Section 105(c) of H.R. 1858 contains a nearly identicalprovision, stating that nothing in the act “shall restrict the rightsof parties freely to enter into licenses or any other contracts withrespect to the use of information.” However, this provision ismade subject to section 106(b) of the bill, which states that aperson shall not be liable for a violation of the act if the personbenefiting from the protection afforded a database under the act“misuses” the protection.188 In determining whether a person hasmisused the protection, a court is to consider, among other factors(1) the extent to which the ability of persons to engage in thepermitted acts under this title has been frustrated by contractualarrangements or technological measures and (2) the extent towhich information contained in a database that is the sole sourceof the information contained therein is made available throughlicensing or sale on reasonable terms and conditions.189 Whilethis provision may not be precisely “comparable” to the EUDirective’s absolute nullification of contract provisions contrary tothe rights of lawful users of databases, it certainly comes closer tothe public policy served by that prohibition than does section 1405of H.R. 354. The upshot is that H.R. 354 may be disqualified as“comparable” to the protection specified by the EU Directiveprecisely because it gives database producers too muchcontractual freedom to vary the terms of the statutory protection.CONCLUSIONIn the short time since the Supreme Court decided Feist,electronic databases have become an integral part of many185. H.R. 354 § 1405(a), 106th Cong. (1999).186. Id. § 1405(b).187. Id. § 1405(e).188. H.R. 1858 §§ 105(c), 106(b), 106th Cong. (1999).189. Id. § 106(b)(1)-(2).

industries. As electronic databases are navigated more efficientlythan printed publications, ensuring incentives for the creation offuture databases is of growing importance to a sound industrialpolicy. Recent technological developments have eroded thenatural lead-time that database developers have heretoforeenjoyed.As we have seen, two distinct legislative responses to Feistcan be found in H.R. 354 and H.R. 1858. Whereas H.R. 354 wouldaccord a database producer rights that are analogous (but notnecessarily “comparable”) to those mandated by the EU DatabaseDirective, thus recreating a limited form of the “sweat of the brow”protection that Feist laid to rest, H.R. 1858 would in effect “refederalize”the holding of International News, at least as it appliesto databases in interstate and foreign commerce, and provideprotection against two specific types of data piracy.The general approach (though not necessarily theenforcement mechanism) of H.R. 1858 is superior to H.R. 354because it prohibits wholesale competitive duplication ofdatabases, while ensuring access to and transformative uses offactual compilations. H.R. 354, by contrast, would extendprotection beyond a database producer’s primary market torelated markets, and would define related markets sufficientlybroadly that a database producer could control a wide range ofnon-competitive uses of data and reserve those potential marketsfor itself.Reincarnations of H.R. 1858 and H.R. 354 are likely to surfaceduring the present congressional session. The passage of eitherbill will have a significant impact on intellectual property lawsand policies in both the United States and worldwide. If theUnited States is perceived to be following the European Union’slead, by according a comparable sui generis right to databasedevelopers, then one would expect similar systems to beimplemented worldwide.However, there is no domestic policy justification for theUnited States to follow the EU’s lead. The economic case forcreating a sui generis right to prevent extraction and reutilizationof unoriginal database contents has never been satisfactorilyexplained by the EU itself. The EU Database Directive franklyseems designed to favor European database producers at theexpense of their customers and non-EU competitors, and topressure the rest of the world to create comparable protection.

The Directive will inevitably drive up the price of databasesproduced in Europe. If the United States resists the temptation toadopt overbroad database legislation, the U.S. database industryshould become all the more competitive in the global market place.Nor will U.S. database producers be all that disadvantaged inEurope itself, as European countries are obliged under Article10bis of the Paris Convention to provide protection against unfaircompetition, which would arguably include protecting against thewholesale competitive duplication of databases created by others.Those U.S. database producers who believe that the new suigeneris protection is crucial for their operations in Europe canqualify for protection by opening a registered office in the EU thatis genuinely linked on an ongoing basis with the economy of anEU member state.If, on the other hand, the United States gives in to thetemptation to adopt overbroad database protection, then U.S.database users (and, in short order, database users worldwide)will have been unwittingly enlisted to subsidize a European effortto gain market share in an industry that the United Statescurrently dominates.

Digital Property: The UltimateBoundary? *Brian F. Fitzgerald **INTRODUCTIONIn a world where information is increasingly relevant to oursocial, cultural and economic existence it is vitally important tounderstand the legal boundary of digital property.1 To this endthis article presents a preliminary framework for understanding*This article was given impetus by an invitation to deliver a Special Lecture at theUniversity of Western Ontario Law School in London Ontario, Canada in October 2000.It was further developed through Spring 2001 while teaching at Santa ClaraUniversity Law School and was further refined through a conference presentation atRoger Williams University School of Law, Bristol, Rhode Island, in April 2001. I owethanks to David Rice, Mark Perry, Margaret Ann Wilkinson, Richard Gold, NicholasPengelley, Michael Lehmann and Anne Fitzgerald.**Brian Fitzgerald is a Professor and Head of the School of Law and Justice atSouthern Cross University in New South Wales, Australia. He is co-editor of one ofAustralia’s leading texts on e-commerce and has published articles on Law and theInternet, Technology Law and Intellectual Property Law in Australia, the UnitedStates, Europe and Japan. Over the past two years, Brian has delivered seminars oninformation technology and intellectual property law in Australia, New Zealand,China, the United States, Canada, Norway and The Netherlands.1. See generally Carl Shapiro & Hal R. Varian, Information Rules: A StrategicGuide to the Network Economy 2 (1999) (explaining strategies for success in networkeconomy); Lester C. Thurow, Building Wealth: New Rules for Individuals, Companiesand Nations in a Knowledge-Based Economy 116-25 (Harper Business ed., 2000)(explaining the importance of intellectual property to modern company’s success orfailure); Frank Webster, Theories of the Information Society (John Urry ed., 1995)(announcing the predominance of information in modern societies); Bernadette E.Lynn, Intellectual Capital: Unearthing Hidden Value by Managing Intellectual Assets,Ivey Bus. J., Jan.-Feb. 2000, at 48 (recognizing the importance of Intellectual Capital);see also Don Tapscott et al., Digital Capital: Harnessing the Power of Business Webs(2000) (discussing the business models used to create wealth in the digital economy);Stephan Bernhut, Measuring the Value of Intellectual Capital, Ivey Bus. J., Mar.-Apr.2001, at 16 (pointing out the vital importance of intellectual capital in the neweconomy).237

the process through which digital property is being constructed.2It aims to highlight through select and recent cases theconstitutional, legislative, contractual and technologicaldimensions of digital property.3I. THE CONSTITUTIONAL MANDATE: THE ENUMERATED INTELLECTUALPROPERTY (IP) POWERIn Australia, Canada and the United States of America, thefederal or national legislature is given significant power in theirrespective federating Constitutions to enact laws with respect tointellectual property. While not all “digital property” comeswithin the constitutional definition of intellectual property4 (asyou will see below), these constitutional sources of the key type ofinformation property, namely, intellectual property (IP), are anobvious and necessary place to begin this analysis. Theconstitutional provisions are expressed as follows:AustraliaSection 51 (xviii) of the Constitution provides that theCommonwealth Parliament may make laws for the peace, orderand good government of the Commonwealth with respect to:“Copyrights, patents of inventions and designs, and trade marks.”CanadaPursuant to sections 91 (22) and 91 (23) of the ConstitutionAct, 1867 the Canadian Parliament has exclusive power tolegislate in respect of “Patents of Invention and Discovery” and2. See also Brian F. Fitzgerald, A Legal Framework for UnderstandingInformational Property Entitlements in the Digital Environment, athttp://www.innovationlaw.org/lawforum/pages/lectureseries.htm (last visited Oct. 20,2001).3. Brian F. Fitzgerald, Intellectual Capital and Law in the Digital Environment,Ivey Bus. J., Mar.-Apr. 2001, at 22.4. See generally Agreement on Trade-Related Aspects of Intellectual PropertyRights, Apr. 15, 1994, Marrakesh Agreement Establishing the World TradeOrganization, Annex 1C, Legal Instruments – Results of the Uruguay Round vol. 31, 33I.L.M. 81 (1994) [hereinafter TRIPS] (defining intellectual property rights). A largeamount of intellectual property legislation is derived from internationalconventions/treaties, with the TRIPS agreement being a current and concise summaryof those international principles.

“Copyrights.” Trademarks are covered by the trade and commercepower.5United StatesArticle I, section 8, clause 8 of the United States Constitutionempowers the Congress to legislate: “To promote the Progress ofScience and useful Arts, by securing for limited Times to Authorsand Inventors the exclusive Right to their respective Writings andDiscoveries.” Trademarks are covered by the Commerce Clause.6Immediately upon reading these provisions one is drawn tocomment that the U.S. provision is written in a more flowery andprosaic language, perhaps representing the style of constitutionaldrafting of an earlier era or one consequent to a revolution. TheU.S. provision appears to be much more explicitly guided by apurpose for granting intellectual property rights—to promote theprogress of science and the useful arts by securing exclusive rightsfor limited times—than are the Australian or Canadianprovisions. The crucial issue that has arisen in relation to thesesources or fonts of intellectual property is the extent to which theyempower governments to propertize information. Are there anylimits? Is it possible to create rights in information that last inperpetuity? Different types of limits are asserted, such as:the purpose for which the IP right is granted/publicdomain user rights;type of information that the constitution will allow to bepropertized;length of time a property right will endure;medium in which the information exists, e.g., fixed in atangible form.The follow up question that has arisen is whether theselimits, if they exist, may be undermined or overridden by otherconstitutional clauses bestowing legislative power at the federallevel or by state laws.7 For example, can the Commerce Clause in5. See McDonald v. Vapor Can. Ltd., [1977] 2 S.C.R. 134 (Can.).6. The Trade-Mark Cases, 100 U.S. 82, 93-95 (1879).7. See Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989);

the United States or the trade and commerce clause in Australiabe used to legislate informational property rights that surpass anyconstitutional limits inherent in the respective IP clauses?8 Orwill the IP clause be read like the unjust acquisition (or takings)clause of the Australian Constitution so as to limit the otherenumerated federal legislative powers?9A. The Rationale and Purpose of IP Rights—The Scope of thePublic DomainA preliminary question and one that is central to the debateover propertizing information is the role and purpose ofintellectual property law or, if looked at from another perspective,the scope of public domain rights.10 Fisher explains that there areKewanee Oil Co. v. Bicron Corp., 416 U.S. 470 (1974); Goldstein v. California, 412 U.S.546 (1973); Compco Corp. v. Day-Brite Lighting, Inc., 376 U.S. 234 (1964); Sears,Roebuck & Co. v. Stiffel Co., 376 U.S. 225 (1964).8. See Brian F. Fitzgerald, (Australian) Constitutional Limits on IntellectualProperty, 23 Eur. Intell. Prop. Rev. 103 (2001).9. See Austl. Const. ch. I, pt. V, § 51, cl. xxxi. The Australian Constitution,amongst other things, divides the legislative powers of the Australian federal systembetween the Commonwealth (central) and State (regional) governments. Section 51 ofthe Constitution provides that the Commonwealth Parliament shall have power tomake laws for the peace, order, and good government of the Commonwealth withrespect to an enumerated list of powers. As a general rule the empowering clauses ofsection 51 are seen as independent sources for legislative action, although section 51(xxxi) is a clear exception to that rule. Section 51 (xxxi) confers a power on theCommonwealth Parliament to legislate with respect to “the acquisition of property onjust terms from any State or person for any purpose in respect of which the Parliamenthas power to make laws.” The provision operates, firstly, to give the Commonwealthpower to acquire property and, secondly, as an individual right or guarantee to privateproperty protected by “just terms.” See Georgiadis v. AOTC (1994) 179 C.L.R. 297(Austl.); Mutual Pools & Staff Ltd. v. Commonwealth (1994) 179 C.L.R. 155 (Austl.). Inacting as a guarantee, the clause is said to remove from other clauses the power tocompulsorily acquire property. However, if another clause in section 51 displays, bywords or content, an intention to allow acquisition without just terms or if theacquisition is of a type that is not susceptible to just terms, the guarantee is notoperative. Therefore statutes imposing taxation, penalties by way of forfeiture ofproperty or a bankruptcy scheme are not seen as infringing the guarantee of or right toprivate property protected by section 51 (xxxi). See Mutual Pools, 179 C.L.R. at 169-71,177-81, 186-89, 219-22; see also Health Insurance Commission v. Peverill (1994) 179C.L.R. 226 (Austl.); Newcrest Mining (WA) Ltd. v. Commonwealth (1997) 190 C.L.R.513 (Austl.).10. See generally Yocahi Benkler, Free as the Air to Common Use: FirstAmendment Constraints on Enclosure of the Public Domain, 74 N.Y.U. L. Rev. 354(1999) (including in the definition of the public domain works protected by copyrightbut that may be used without the permission of the copyright owner under privilegessuch as the fair use doctrine); Jessica Litman, The Public Domain, 39 Emory L.J. 965

(at least) four (imperfect yet useful) overarching theories ofintellectual property.11Economic/utilitarian theory: intellectual property law isjustified in terms of economic efficiency.12Lockean/labor desert: intellectual property rights arenatural rights earned through adding labor to thecommon resource of information with the proviso thatenough and as good is left for others.13Personhood: intellectual property is an emanation of theperson and the law should facilitate this personalaspect.14Social planning: intellectual property law should bedesigned to culturally enrich democratic society.15Economic RightsArguably, the guiding premise of American copyright and(1990) (defining the public domain as the realm unprotected by copyright).11. William Fisher, Theories of Intellectual Property, in New Essays in the Legaland Political Theory of Property, 169-73 (Stephen R. Munzer ed., 2001); see also AnneFitzgerald, Intellectual Property 7-9 (Law Book Co., Sydney 1999); Robert P. Merges etal., Intellectual Property in the New Technological Age 1-21 (2d ed. 2000); JustinHughes, The Philosophy of Intellectual Property, 77 Geo. L.J. 287, 288-89 (1988).12. See, e.g., Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 429-32 (1984); Millar v. Taylor, 98 Eng. Rep. 201, 218, 252-53 (K.B. 1769); Adam Smith,Lectures on Jurisprudence 83 (Ronald L. Meek et al. eds., Liberty Classics 1982)(1762); William M. Landes & Richard A. Posner, An Economic Analysis of CopyrightLaw, 18 J. Legal Stud. 325 (1989); Neil W. Netanel, Copyright and a Democratic CivilSociety, 106 Yale L.J. 283 (1996).13. See Universal Declaration of Human Rights, G.A. Res. 217 A (III), U.N. GAOR3d Sess., pt. 1, at art. 27(2), U.N. Doc. A/RES/217 A (III) (1948) (stating “[e]veryone hasthe right to the protection of the moral and material interests resulting from anyscientific, literary or artistic production of which he is the author.”); Wendy J. Gordon,A Property Right in Self Expression: Equality and Individualism in the Natural Law ofIntellectual Property, 102 Yale L.J. 1533 (1993).14. See generally Margaret J. Radin, Property and Personhood, 34 Stan. L. Rev.957 (1982) (exploring the relationship between property and personhood).15. See generally Peter Jaszi, Toward a Theory of Copyright: The Metamorphosesof “Authorship”, 1991 Duke L.J. 455 (1991); The Construction of Authorship: TextualAppropriation in Law and Literature (Martha Woodmansee & Peter Jaszi eds., 1994)(discussing the relationship between copyright and literary and artistic culture).

patent law is the utilitarian ethic that legal protection ofintellectual property is needed to advance public welfare becauseit fosters creative genius/product, which can in turn be distributedfor the good of the general public.16 In the words of the UnitedStates Supreme Court in Mazer v. Stein:17The economic philosophy behind the clause empoweringCongress to grant patents and copyrights is the conviction thatencouragement of individual effort by personal gain is the bestway to advance public welfare through the talents of authors andinventors in ‘Science and useful Arts.’18This economic conception of intellectual property is reinforcedby the copyright and patent statutes, which bestow rights uponthe IP owner to control the economic exploitation of the work orproduct. The intellectual property law of the United Kingdom,Canada and Australia has exhibited a similar utilitarian economicbasis.19 As well, in these systems the IP owner holds economicrights to exploit the information, e.g., the exclusive right of thecopyright owner to control reproduction.Lockean Natural RightsThe view that a person possesses a natural right to own whatthey have created through their labour is also very influential inmany intellectual property systems throughout the world.20 Inmore recent times, though, copyright claims to own information onthe basis of labour have not been accepted in the United States orCanada per se, although a recent decision in Australia was morereceptive to such claims.21 If copyright is to be the vehicle for theclaim to own the value of adding one’s labour to the common stockof information in the United States and Canada, then the labourmust produce an intellectual product. Otherwise mere “sweat of16. See e.g., Paul Goldstein, Copyright’s Highway 165-96 (1994).17. 347 U.S. 201 (1953).18. Id. at 219.19. See Welcome Real-Time SA v. Catuity Inc. [2001] F.C.A. 445, at para. 129(Austl.); David Vaver, Intellectual Property Law 6-13 (1997); David Fewer,Constitutionalizing Copyright: Freedom of Expression and the Limits of Copyright inCanada, 55 U.T. Fac. L. Rev. 175, 187-93 (1997).20. See Charter of Fundamental Rights of the European Union art. 17, at para. 2.(“Intellectual property shall be protected.”); Fewer, supra note 19, at 187-89, 191-93(discussing the application of this theory to Canadian intellectual property law).21. See Telstra Corp. v. Desktop Marketing Sys. Pty. Ltd. [2001] F.C.A. 612(Austl.).

the brow” claims in those countries will need to be asserted undersui generis law like the emerging European database laws, unfaircompetition law22 or unjust enrichment law.23Moral RightsIn contrast to the Anglo-American approach, the intellectualproperty law of continental Europe (especially copyright law) haslong recognised both an economic and personal aspect tointellectual property rights.24 The personal aspect which isencompassed by the third category of theories in Fisher’staxonomy is known as a moral right.25Moral rights are personal rights belonging to authors orcreators of copyright material and exist quite independently fromeconomic rights. They continue to exist even after the economicrights have been transferred. The principal moral rights are:The right of attribution, that is the right of the creator of22. See generally Int’l News Serv. v. Associated Press, 248 U.S. 215 (1918)(discussing unfair competition law); Nat’l Basketball Ass’n v. Motorola, Inc., 105 F.3d841 (2d Cir. 1997) (ruling professional basketball games were not “original works ofauthorship” protected by Copyright Act); Bd. of Trade of City of Chicago v. Dow Jones,456 N.E. 2d 84 (Ill. 1983) (ruling publisher’s stock market indexes and averages couldnot be used by board of trade as the basis for its proposed stock index futures contractswithout the consent of the publisher); Wendy J. Gordon, On Owning Information:Intellectual Property and a Restitutionary Impulse, 78 Va. L. Rev. 149, 281 (1992)(suggesting that “the judicial experience in the allied area of restitution has suggestedthat economic and other norms should, and do, condition the implementation of theimpulse to grant reward for labor expended”); J.H. Reichman & Pamela Samuelson,Intellectual Property Rights in Data?, 50 Vand. L. Rev. 51, 56 (1997) (proposing “eitherthe use of unfair competition principles to protect database contents, or the adoption ofan intellectual property regime based on more refined liability principles, rather thanon exclusive property rights, that would reconcile the need for legal incentives to investwith a calculus of net social benefits”).23. See Matarese v. Moore-McCormack Lines, Inc., 158 F.2d 631 (2d Cir. 1946);Bristol v. Equitable Life Assurance Soc’y of U.S., 30 N.E. 506 (N.Y. 1892); Brian F.Fitzgerald & Leif Gamertsfelder, Protecting Informational Products (IncludingDatabases) Through Unjust Enrichment Law: An Australian Perspective, 20 Eur. Intell.Prop. Rev. 244 (1998).24. See Joined Cases C-241-242/91, Radio Telefis Eireann and Indep. TelevisionPubl’ns Ltd. v. Comm’n, 1995 E.C.R. I-743, at para. 71, [1995] 4 C.M.L.R. 718 (1995)(explaining “The Court of First Instance is right . . . in stating that the essentialfunction of copyright is to protect the moral rights in the work and ensure a reward forcreative effort”); see also Universal Declaration of Human Rights, supra note 15; seegenerally Millar, 98 Eng. Rep. at 252-53; Gerald Dworkin, Moral Rights and theCommon Law Countries, 5 Austl. Intell. Prop. L.J. 5 (1994).25. Fisher, supra note 12.

a work to be publicly identified as such and to preventothers from claiming authorship of the work, to preventothers from wrongfully attributing to an author worksthat are not his or hers or that are unauthorised alteredversions of his or her work; andThe right of integrity, that is the right to object todistortions or derogatory distortions of a work.26The obligation to afford moral rights to creators of copyrightmaterials arises from Article 6bis of the Berne Convention, whichprovides that:(1) Independently of the author’s economic rights, andeven after the transfer of the said rights, the author shallhave the right to claim authorship of the work and toobject to any distortion, mutilation or other modificationof, or other derogatory action in relation to, the said work,which would be prejudicial to his honour or reputation.(2) The rights granted to the author in accordance withthe preceding paragraph shall, after his death, bemaintained, at least until the expiry of the economicrights, and shall be exercisable by the persons orinstitutions authorised by the legislation of the countrywhere protection is claimed. However, those countrieswhose legislation, at the moment of their ratification of oraccession to this Act, does not provide for the protectionafter the death of the author of all the rights set out inthe preceding paragraph may provide that some of theserights may, after his death, cease to be maintained.(3) The means of redress for safeguarding the rightsgranted by this Article shall be governed by thelegislation of the country where protection is claimed.Until recently, Australian law provided only minimal andindirect recognition for moral rights, notwithstanding Article 6bis ofthe Berne Convention. The recognition of moral rights under theCopyright Act, 1968, was limited to a duty not to falsely attributethe authorship of a work, a duty not to falsely attribute the26. Brian F. Fitzgerald & Anne Fitzgerald, Cyberlaw: Cases and Materials on theInternet, Digital Intellectual Property and Electronic Commerce 336 (Prospect Media,Sydney 2002).

authorship of an altered work and a duty not to falsely attribute theauthorship of a reproduction of an artistic work. The CopyrightAmendment (Moral Rights) Act, 2000, which entered into force onDecember 21, 2000, introduces the moral rights of integrity andattribution of authorship and strengthens the earlier prohibitionon false attribution of authorship. Under the new Australian law,moral rights are not infringed by acts to which the creator hasconsented, providing the consent is in writing and meets certainother conditions.Interestingly, Canada introduced moral rights in theCopyright Act of 1931; however, scope for waiver has meant that aculture of moral rights has not been prominent in Canadiancopyright law.27 The United States has legislated for some verylimited applications of moral rights,28 however many Americanacademics are wary of the potential of moral rights to lock upinformation and inhibit free speech, especially in the context ofthe Internet.29Culturally Enriching RightsIn a world where digitized information (and itscontrol/ownership) is of continuing and increasing significance tothe economy, society and culture, we have seen renewed relevancefor and interest in Fisher’s fourth category of IP theory: socialplanning or cultural enhancement theory. Every significant claimover digital value at issue at the moment can be described as oneinvolving a battle between utilitarian or Lockean/natural rightsclaims to ownership versus the creation of intellectual propertyrights for reasons of cultural enhancement. In the digitalenvironment of an information society, where “information islifeblood,” the cultural aspects of intellectual property are broughtinto focus. The decentering of the author as creator is part of thisphenomenon, but more so the realisation that an expressive workis a collection of culturally imbued facts. As we shall see below, in27. See Vaver, supra note 19, at 87-96.28. See generally Visual Artists Rights Act, 17 U.S.C. § 106A (1994 & Supp. IV1998) (granting rights of attribution and integrity to certain authors of visual art);Carter v. Helmsley-Spear, Inc., 71 F.3d 77 (2d Cir. 1995) (summarizing the scope ofmoral rights in the law of the United States).29. See generally William W. Fisher III, Property and Contract on the Internet, 73Chi.-Kent L. Rev. 1203 (1998) (suggesting moral rights should have more limitedapplication in the eclectic transnational and digitized world of the Internet).

the United States, the power of the First Amendment is beingemployed to attack the “private property” imperialism of the IPstatutes over information and to thereby assist the case of theculture theorists.If we live in an information based economy, culture andsociety, the process of propertizing information30 must be seen asbeing inherently concerned with the way we live, think,communicate and construct knowledge. This takes us far beyondquestions of economics to key cultural and social issues, which theprocess of propertizing information must now accommodate. Forinstance, it is no longer good enough to allocate property rights insoftware without considering the social and cultural implicationsof the market power and monopoly of thought that such a processwill engender. Software is not just code, it is discourse, in that itallows us to see and say things in digital space.31Free software and open code software projects, in someinstances, epitomize a cultural rationale for intellectual propertyrights.32 In the classic free software scenario embodied in the30. See Felix S. Cohen, Transcendental Nonsense and the Functional Approach, 35Colum. L. Rev. 809, 814-17 (1935); see e.g., E. Richard Gold, Body Parts: ProprietaryRights and the Ownership of Human Biological Materials (Geo. Univ. Press 1996);Ejan Mackaay, The Economics of Emergent Property Rights on the Internet, in TheFuture of Copyright in a Digital Environment 13 (P. Bernt Hugenholtz ed., 1996)(stating that although traditional property fencing techniques do not work well on theInternet, property is by no means dead, because of the initiatives exploring alternativefencing and new extensions of property rights); Thomas A. Stewart, IntellectualCapital: The New Wealth of Organizations (Bantam Doubleday Dell 1997); JamesBoyle, Cruel, Mean or Lavish? Economic Analysis, Price Discrimination and DigitalIntellectual Property, 53 Vand. L. Rev. 2007 (2000); Julie E. Cohen, Copyright and thePerfect Curve, 53 Vand. L. Rev. 1799 (2000); Michael A. Heller, The Tragedy of theAnticommons: Property in the Transition From Marx to Markets, 11 Harv. L. Rev. 622(1998) (discussing the dangers of anticommons property, specifically the under use ofresources and the lack of effective privileges to use); David Lange, Recognizing ThePublic Domain, Law & Contemp. Probs., Autumn 1981, at 147; Pamela Samuelson &Kurt Opsahl, Licensing Information in the Global Information Market: Freedom ofContract Meets Public Policy, 21 Eur. Intell. Prop. Rev. 387 (1999); Lester C. Thurow,Needed: A New System of Intellectual Property Rights, Harv. Bus. Rev., Sept.-Oct. 1997,at 95; Hal R. Varian, Versioning Information Goods, (Mar. 13, 1997) (discussingdifferential pricing known as quality discrimination or versioning), athttp://www.sims.berkeley.edu/~hal/people/hal/papers.html.31. See Brian F. Fitzgerald, Software as Discourse: The Power of IntellectualProperty in Digital Architecture, 18 Cardozo Arts & Ent. L.J. 337 (2000).32. See Marcus Maher, Open Source Software: The Success of an AlternativeIntellectual Property Incentive Paradigm, 10 Fordham Intell. Prop. Media & Ent. L.J.619 (2000) (showing how the science of complexity theory is able to explain the opensource movement’s ability to translate non-economic incentive mechanisms into a

GNU General Public Licence (GPL), code is distributed in amanner that is open and free (not free as in beer but free as inspeech), allowing software developers further down the line tomodify and improve upon the initial software product. The initialdistributor of the code controls its presentation and furtherdissemination through copyright and contract law (contractualsoftware license). In general, subsequent users and modifiers arerequired to further distribute the code they received and to makecode of any derivative work available to the public. In thisprocess, copyright law is used to create a “copyleft” effect asopposed to a “copyright” effect by mandating that code should beopen and free for all to use in the innovation and development ofsoftware. In doing so, copyright law is used to structure thepractices of software developers into a community of effort, whilein a broader sense ensuring diversity and social innovation of thedigital environment and, thereby, social existence. In short, freesoftware and open code projects have the potential to useintellectual property law (copyright) to enrich and to betterstructure cultural existence. Not every developer or all projectsunder these banners will be minded to act in a culturallybenevolent manner—some will do it for the purpose of building areputation that can later be exploited for monetary rewards andothers will increasingly use open code to augment already thrivingproprietary business models.33Two difficult issues remain to be considered in relation to theculture theory. Firstly, how does it relate to the first theoryconsidered, which is primarily focused on public welfare in aneconomic sense? There can be little doubt that both theory oneand four are concerned with public welfare, but in different ways.process for technological development and innovation); David McGowan, LegalImplications of Open-Source Software, 2001 U. Ill. L. Rev. 1, (2001); Stephen M.McJohn, The Paradoxes of Free Software, 9 Geo. Mason L. Rev. 25 (2000) (concludingthat open source software may have a greater effect on the law of developingtechnologies than the law will have on software practices); David Bollier, The Power ofOpenness: Why Citizens, Education, Government and Business Should Care About theComing Revolution in Open Source Code Software, available athttp://eon.law.harvard.edu/opencode/h2o/#intro (last visited Oct. 27, 2001) (illustratingthe broad effect of changes in open code software); www.gnu.org; www.opensource.org(last visited Oct. 27, 2001).33. See generally Eric S. Raymond, The Cathedral and the Bazaar, available athttp://www.tuxedo.org/~esr/writings/cathedral-bazaar/cathedral-bazaar/index.html(last visited Oct. 27, 2001) (discussing the evolution of Linux in the commercial world).

Theory one proposes to reward the author in the name of publicwelfare, but the benefit to community or culture becomes a“presumed” consequence of the creative activity.34 Theory four isthe reverse in that it starts out with the community or culture andseeks to explain when it is beneficial for a creator to become aproperty holder. In this equation, economic efficiency may beconsidered but is by no means a totalising criterion or rationale.35The conclusion must be that both seek to prosper public welfarebut from completely different ends of the spectrum.Secondly, claims for protection of indigenous culturalmaterials are also subject to increasing attention. These might beseen as claims to collective or group moral rights or, more broadly,as claims that IP law should be used to respect and enhancecultural tradition and spirituality.36 A concern is that grantingperpetual rights to control information that has special culturalsignificance to indigenous peoples locks up information in asociety where access to and use of information should beoptimised. Another concern is that if indigenous peoples can beallowed to claim the special cultural significance of information,then a way will be opened for non-indigenous people to makesimilar claims in respect of other (religious or social) informationculturally significant to them. These responses understate therichness of indigenous culture and its plight in the face ofunethical economic exploitation.Proprietary Purpose: A Limiting FactorWhether intellectual property rights are designed to rewardpeople, or to reduce transaction costs and increase allocativeefficiency, or for reasons of rewarding the fruits of labour, orpersonhood theories or cultural development, the question34. See Benkler, supra note 10; Boyle, supra note 30; Julie E. Cohen, supra note30, (criticising the presumption that Intellectual Property law is always economicallyand culturally beneficial).35. It is interesting to contemplate how a cultural approach to IP might bereconciled with the permitted exceptions stipulated under TRIPS, arts. 13, 17, 30.36. Maroochy Barambah & Ade Kukoyi, Protocols for the Use of IndigenousCultural Material, in Going Digital 2000: Legal Issues for E-Commerce Software andthe Internet 133 (Prospect Publishing Sydney ed., 2000); Terri Janke, Our Culture OurFuture: A Report on Australian Indigenous Cultural Intellectual Property Rights, 43-48 (1988) (stating indigenous people’s desires for protection of indigenous cultural andintellectual property); United Nations Draft Declaration on the Rights of IndigenousPeoples, art 29 (1993).

emains whether any purpose or rationale for such rights acts as alimit on what rights can be legislated?There is sufficient United States Supreme Courtjurisprudence to suggest that copyright or patent legislationshould stimulate economic efficiency. But in light of the recentdecision in Eldred v. Reno,37 it might be doubted whether aUnited States court would use this as the basis for any generaldecision limiting the power of Congress to legislate. If anything,such a rationale for the IP clause would most likely beimplemented through interpretation of the various elements of theclause. Notions of personhood38 or cultural enhancement do notseem to be constitutionally required or entrenched. However, thefollowing statement from Graham v. John Deere Co,39 read at itsbroadest, suggests that there may be a broader purposive limitwhich, with some imagination, could embrace notions of socialimpact:The Congress in the exercise of the patent power may notoverreach the restraints imposed by the stated constitutionalpurpose. Nor may it enlarge the patent monopoly without regardto the innovation, advancement or social benefit gained thereby.40In Australia and Canada it is doubtful whether any generallimit based on notions of economy, personhood or culture would beoperational. Once again, if any limits exist they must be found inthe specific elements of the respective constitutional clauses.Although, the recent statement by Justice Kirby of the High Courtof Australia in Grain Pool of WA v. The Commonwealth41 suggestsdemocratic and cultural considerations may be relevant tointerpreting the scope of legislative power on this issue:No absolute or unlimited rule may be stated. The protection37. 239 F.3d 372 (D.C. Cir. 2001). In Eldred, the majority held that theperambulatory words of the United States Copyright and Patent Clause—to promoteprogress of science and the useful arts—are not to be read as limits on the legislativepower of Congress. This suggests that the purpose for bestowing the legislative powerbecomes relevant if at all in relation to the operative parts or terms of the clause.38. See Gilliam v. ABC Inc., 538 F.2d 14 (2d Cir. 1976). These issues can berelevant if related to the economic incentive argument.39. 383 U.S. 1, 5-6 (1966).40. Id.; see also Sears, Roebuck & Co., 376 U.S., at 230 (quoting Kendall v.Winsor, 21 How. 322, 328 (1859)) (“[I]n rewarding useful invention, the ‘rights andwelfare of the community must be fairly dealt with and effectually guarded.’”).41. [2000] H.C.A. 14, available athttp://www.austlii.edu.au/au/cases/cth/high_ct/2000/14.html.

of intellectual property rights must be afforded in a constitutionalsetting which upholds other values of public good in arepresentative democracy. In the United States the relevant headof constitutional power has been viewed as containing in-builtlimitations many of which are derived from the competingconstitutional objective of public access to information: Graham v.John Deere Co.; Feist Publications, Inc v. Rural Telephone ServiceCo., Inc. In Australia, the constitutional setting is different butthe existence of competing constitutional objectives, express andimplied, is undoubted.42Having considered the difficulty courts have with invoking abroad notion of purpose for granting IP rights as an enforceablerestraint on legislation, the more interesting question is todetermine the specific limits that arise from the elements of theclauses.B. The Type of Information That the Constitution Will Allow to bePropertizedCentral to the presentation of digitised information issoftware. Software has been protected as a literary work underthe US Copyright Act since 1980 and under the AustralianCopyright Act since 1984. Article 10 (1) of the TRIPs agreementalso provides that software shall be protected as a literary text incopyright law.43 More recently, software has been subject to avast amount of patenting throughout the world.44 It is clear thenthat the most prominent form of digital property has been held tobe a type of information that can be subject to IP rights. However,not all information, including digitised information, isconstitutionally eligible for protection under the copyright andpatent heads of power.1. Raw DataThe classic example of unprotected information is raw data or42. Id., at n.218 (citations omitted). Judge Kirby also talks of “potential benefit tothe community.”; see also Lawrence Lessig, Code and Other Laws of Cyberspace 131,133-34 (Basic Books 1999); Fitzgerald, supra note 8.43. TRIPS, supra note 5, at art. 10 cl. 2 (providing that “Computer programs,whether in source or object code, shall be protected as literary works under the BerneConvention (1971)”).44. See, e.g., State St. Bank & Trust Co. v. Signature Fin. Group, Inc., 149 F. 3d.1368 (Fed. Cir. 1998); Welcome Real-Time, [2001] F.C.A. 445.

ideas. Copyright law has long taken the view that it protects theexpression of data/ideas but not the raw data/ideas themselves.45This view is known as the idea/expression dichotomy and is nowembodied in the TRIPs agreement: “Copyright protection shallextend to expressions and not to ideas, procedures, methods ofoperation or mathematical concepts as such.”46In the United States Supreme Court decision in FeistPublications, Inc. v. Rural Telephone Service Co.,47 it was heldthat raw data, namely entries in a telephone directory, was notsubject to copyright protection. The Court explained:The sine qua non of copyright is originality. To qualifyfor copyright protection, a work must be original to theauthor . . . . Original, as the term is used in copyright,means only that the work was independently created bythe author (as opposed to copied from other works), andthat it possesses at least some minimal degree ofcreativity. To be sure, the requisite level of creativity isextremely low; even a slight amount will suffice. Thevast majority of works make the grade quite easily, asthey possess some creative spark, “no matter how crude,humble or obvious” it might be . . . . Originality does notsignify novelty; a work may be original even though itclosely resembles other works, so long as the similarity isfortuitous, not the result of copying. To illustrate,assume that two poets, each ignorant of the other,compose identical poems. Neither work is novel, yet bothare original and, hence, copyrightable.Originality is a constitutional requirement . . . In TheTrade-Mark Cases, the Court addressed theconstitutional scope of “writings.” For a particular workto be classified “under the head of writings of authors,”the Court determined, “originality is required.” TheCourt explained that originality requires independentcreation plus a modicum of creativity: “[W]hile the wordwritings may be liberally construed, as it has been, toinclude original designs for engraving, prints, &c., it is45. See Feist Publ’ns, Inc. v. Rural Telephone Serv. Co., 499 U.S. 340 (1991).46. TRIPS, supra note 5, at art. 9 cl. 2.47. 499 U.S. 340 (1991).

only such as are original, and are founded in the creativepowers of the mind. The writings which are to beprotected are the fruits of intellectual labor, embodied inthe form of books, prints, engravings, and the like.”In Burrow-Giles, the Court distilled the samerequirement from the Constitution’s use of the word“authors.” The Court defined “author,” in a constitutionalsense, to mean “he to whom anything owes its origin;originator; maker.” As in The Trade-Mark Cases, theCourt emphasized the creative component of originality.It described copyright as being limited to “originalintellectual conceptions of the author,” and stressed theimportance of requiring an author who accuses another ofinfringement to prove “the existence of those facts oforiginality, of intellectual production, of thought, andconception.”“No one may claim originality as to facts.” This isbecause facts do not owe their origin to an act ofauthorship. The distinction is one between creation anddiscovery: the first person to find and report a particularfact has not created the fact; he or she has merelydiscovered its existence. To borrow from Burrow-Giles,one who discovers a fact is not its “maker” or“originator.”48Building upon this reasoning, the Court held that the copyingof 1309 entries in the white pages of the telephone directory wasnot a copyright infringement nor was the selection or arrangementof the entries in basic alphabetical order creative enough to give athin layer of copyright protection to the arrangement of theentries.49As a consequence of the Feist decision, arguments were madethroughout the world for the creation of a sui generis form of legalprotection for the time and expense of creating databases. In1996, the European Union adopted a Directive on the LegalProtection of Databases50 that proposed a sui generis regime; later48. Id. at 345-47 (citations omitted).49. See BellSouth Advertising & Pub. Co. v. Donnelley Info. Pub., Inc., 999 F.2d1436 (11th Cir. 1993).50. Council Directive 96/9/EC, 1996 O.J. (L77) 20.

in 1996, the World Intellectual Property Organisation (WIPO)produced proposals for a similar international scheme for theprotection of databases. The EU Directive creates an exclusive suigeneris right for the makers of databases.51 The general objectiveof this right is to protect the investment of time, money and effortby the maker of a database, irrespective of whether the databaseis in itself innovative. According to the Directive, a database isprotected if there has been a substantial qualitative orquantitative investment in obtaining, verifying or presenting thecontents of the database.52 The duration of the protectionprovided by the Directive is fifteen years.53At present, the WIPO proposal has stalled. However, the EUDirective is slowly being introduced.54 The U.S. Congress hasconsidered a number of proposals for database protection but hasyet to enact a definitive sui generis regime. One proposal from1999 was the Collections of Information Antipiracy Act,55 whichread in part:1402. Prohibition against misappropriationAny person who extracts, or uses in commerce, all or asubstantial part, measured either quantitatively orqualitatively, of a collection of information gathered,organized, or maintained by another person through theinvestment of substantial monetary or other resources, soas to cause harm to the actual or potential market of thatother person, or a successor in interest of that otherperson, for a product or service that incorporates thatcollection of information and is offered or intended to beoffered for sale or otherwise in commerce by that otherperson, or a successor in interest of that person, shall beliable to that person or successor in interest for the51. Id. at art. 8. The reciprocity principle embodied in art. 11 means thatdatabase makers from countries outside the EU will not be given the benefits of thesedatabase rights unless their countries offer comparable protection to EU databasemakers.52. See id. at art. 7.53. Id. at art. 10.54. See, e.g., Council Directive 96/9/EC, 1996 O.J. (L77) 20, implemented by theCopyright and Rights in Databases Regulations 1997 (SI 1997/3032) (implementing theEU directive in the UK on Jan. 1, 1998); British Horseracing Board Ltd. v. William HillOrg., [2001] R.P.C. 31 (Ch. 2001) (Eng.).55. H.R. 354, 106th Cong. (1999).

emedies set forth in section 1406.This Act proposed to protect databases for up to fifteen years.56 Amore “user friendly” proposal was the Consumer and InvestorAccess to Information Act.57 Variations of both proposals are stillbeing considered.58 Canada and Australia do not have a suigeneris regime for database protection although both countrieshave been examining the issue for some time.An interesting question is whether the Commerce Clausecould be used to legislate such sui generis database rights inAustralia or the United States. This issue is dealt with more fullybelow; suffice to say, at this point the answer depends on whetherthe IP clause is seen to guarantee an individual’s right to accessraw data, override other federal constitutional clauses.While the Canadian59 and Australian courts may not giveexactly the same definition to the terms, there could be littleconjecture that “copyright,” as listed in the respectiveconstitutions, embodies notions of independent creation and amodicum of creativity.60 However, a recent case in Australiaconcerning telephone directories has suggested that“compilations” may not require intellectual input (creativity) tosecure copyright protection.In Telstra Corporation Limited v. Desktop Marketing SystemsPty. Ltd.,61 Judge Finkelstein of the Federal Court of Australiaheld that Australia’s leading telecommunications provider,Telstra, could maintain an action for copyright infringement basedon their white and yellow pages telephone directories.62 In that56. Id. § 1408.57. H.R. 1858, 106th Cong. (1999).58. See Reichman & Samuelson, supra note 22; see also, Charles R. McManis &Christopher R. Alder, Database Protection in the Digital Information Age, 7 RogerWilliams Univ. L. Rev. 9 (2001).59. See Tele-Direct (Publications) Inc. v. Am. Bus. Info. Inc., [1997] 154 D.L.R. 4th328 (Can.).60. See e.g., Kalamazoo (Austl.) Pty. Ltd. v. Compact Bus. Sys. Pty. Ltd. (1985) 5I.P.R. 213 (Austl.); ITP Pty. Ltd. v. United Capital Pty. Ltd. (1985) 5 I.P.R. 315 (Austl.);Ladbroke (Football) Ltd. v. William Hill (Football) Ltd., [1964] 1 W.L.R. 273 (Eng.); cf.G A Cramp & Sons Ltd. v. Frank Smythson Ltd. [1944] A.C. 329 (Austl.); Vaver, supranote 19, at 41-45; see also Copyright Law Review Committee, Report on theSimplification of the Copyright Act, 1968, Part 2, Categorisation of Subject Matter andExclusive Rights and Other Issues, paras. 5.42-.47 (1999) (noting their views on theproduction of copyright material with computers), at http://www.agps.gov.au/clrc.61. [2001] F.C.A. 612.62. Id. at para. 147.

case, the defendants had produced CD versions of the books with anumber of value added features such as allowing people to searchfor names through postcode, address or reverse phone numbersearching.63 Judge Finkelstein, relying on a series of Englishdecisions, held that in the case of compilations it was notnecessary for there to be originality in the sense of intellectualinput or creativity; sweat of the brow was sufficient.64 Heexplained:There is debate as to the meaning of “originality” incopyright law. Everyone agrees that for a work to be“original” it must be independently created by the author;that is, it must not be copied from another work. Butthere is a further requirement. The disagreementconcerns the nature of that requirement. There are twoschools of thought. On the one hand there are those whosay that copyright will only protect intellectual effort, andunless there is at least some intellectual labour in thecreation of a work, it cannot be “original”. . . .On the other hand there is a school that argues that atleast in respect of a compilation, the originalityrequirement will be satisfied if there has been some effortexpended in producing the work, especially effort ingathering or collecting the factual data that is63. Id. at para. 24.The primary data that is used to produce DtMS’ products is taken fromTelstra’s white pages directories and yellow pages directories. A Sydneycompany, Dependable Database Data Pty. Ltd., sends every white pagesdirectory and yellow pages directory to the Philippines where a large teamtypes (the technical term is key punches) all the listings into a computer.The information is changed in various respects. For example, abbreviationsare expanded (Rd to Road), the name of the State is included, and theinformation is formatted so that it may easily be read by a computer. Theinformation is then provided to DtMS in electronic form, usually as acomputer disk.The information is loaded onto the computer system maintained byDtMS. There is some cursory validation of the data, but not to any significantextent. Then information is added, such as the appropriate industry code,number of employees, facsimile number and Ausdoc DX number. Theinformation is updated if errors are identified, or if there is a “do not contact”attached to a particular nameId. at paras. 25-26.64. Id. at paras. 84-85.

eproduced, though there be no ingenuity in thearrangement or presentation of that data.Quite apart from matters of policy, it might be said thatthe very nature of copyright requires the work to be theproduct of creative thought. The first copyright statutewas enacted “for the Encouragement of learned Men tocompose and write useful Books.” These books werenecessarily the result of the author’s intellectual effort.That was also true of other works (engravings,sculptures, dramatic works and the like) that were givencopyright protection by the early statutes. But acompilation is of a different character from a work of artor literature. This is especially true of a compilation offacts that are in the public domain. For this type ofcompilation to come into existence, the facts must beselected, collected, arranged in a particular fashion, andthen produced in some form. It may be possible todescribe as creative the processes of selection, collectionor arrangement. But the creativity is of a different orderfrom that involved in producing a work of art orliterature. The English cases seem implicitly to acceptthis proposition, because the originality requirement for acompilation could not be the same as for other works.More particularly, when it was held that copyright couldsubsist in a work such as a chemist’s stock list (Collis v.Cater), the preparation of which involved no intellectualeffort, it was apparent that a work could be original inthe absence of creativity. The old view of originality haddisappeared. So, copyright protection could be claimed bya person who brought out a directory in consequence of anexpensive, complicated and well organised venture, evenif there was no creativity in the selection or arrangementof the data.That being the law in England in 1911, it became thelaw in Australia in 1912 when the English statute wasadopted as the law in this country.65The difficulty with the approach espoused by Judge Finkelstein is65. Id. at paras. 8, 9, 84-85.

that if creativity is not required as an element of originality thencopyright is awarded for the effort of collection and, in essence,provides copyright protection of the facts. This approach becomesproblematic if the judge, (as Judge Finkelstein did in this case),takes a broad view of what amounts to copying in order to protectthe labour of the data base compiler. While the CD versions of thebooks contained enhancements, the judge seem minded to protectthe labours of Telstra from wholesale copying by finding thatinfringement had occurred:In substance, the respondents say that “the look, feel,arrangement, functionality, extent of information,accuracy and purpose [of the CD-roms] are all radicallydifferent” from Telstra’s products. They also contend thatthe visual appearance is different because the entries arealphabetically ordered by postcode, not arranged ingeographical areas, contain full mailing addresses, have alayout that is continuous rather than in columns onpages, and are displayed in different fonts. Whencompared with the yellow pages directories, therespondents say that these differences are heightened bythe fact that no advertisements are taken and theheadings are not proximate to a group of business namesand addresses. Instead, each business entry has itsclassification entered as part of the data shown inrelation to that business. As for the headings, it is saidthat there has been no reproduction of the compilation ofheadings because what has been reproduced areindividual headings in relation to individual records ofdata . . . .In this case, the substance of the information that hasbeen taken from Telstra’s works (the directory portion ofthe directories and the headings that appear in the yellowpages directories and headings books) has beenreproduced in the CD-roms. It must be remembered thatcopyright is not claimed for each particular entry,because copyright does not subsist in each individualrecorded fact. It is claimed in the whole of the collecteddata, ordered in a particular way. As regards thedirectories, the significant recorded facts (name, address,telephone number, and the relevant type of business) are

the same, or substantially the same, as they appear inTelstra’s works. While there are differences, they are inthe detail. For example, when displayed on a screen, theinformation from the CD-rom does not appear as columnson a page. But the information can be retrieved inalphabetical order (by postcode rather than region) andcan be examined in much the same way as one wouldread a column on a page. The fact that the alphabeticallistings are by postcode and not region, is not a materialdifference. Nor is the fact that portions of theadvertisements are not reproduced. As regards theheadings, it is true that they appear once only in eachyellow pages directory and that the heading appears witheach business entry in the CD-roms. This difference isimmaterial. All the headings have been taken, as haveall the listings beneath those headings. The appearanceof the headings and the listings in the CD-roms issufficiently similar to constitute a reproduction.66The fear becomes that copyright law then takes on the job ofprotecting the compilation of data without adequate safeguardsagainst locking up information. With the emergence of sui generisdatabase regimes in Europe and, inevitably, the United States, itwill be interesting to see how that case is decided if it goes to ahigher court on appeal.2. Plant Breeder’s Rights—Are They Patents?For digital property to be eligible for legislative constructionand protection under the respective IP clauses, it must fall withinsome general notion of copyright or patent (and in Australia,design or trademark). A recent case on plant breeder’s rightsshows that, at least in Australia, the High Court is reluctant toconstrue the constitutional mandate too narrowly so as to excludeinnovative informational products.In Grain Pool of WA v. Commonwealth,67 the plaintiff claimedthat two Acts of the Parliament of the Commonwealth of Australiawere beyond the enumerated legislative powers of thatparliament. The High Court of Australia was asked to considerwhether the two pieces of legislation—the Plant Variety Rights66. Id. at paras. 104, 109.67. [2000] H.C.A. 14.

Act, 1987 (“the Varieties Act”) and its successor, the PlantBreeder’s Rights Act, 1994 (“the Breeder’s Rights Act”)—were dulyenacted pursuant to section 51 (xviii) of the Constitution, asexplained above, provides that the Commonwealth Parliamentmay make laws for the peace, order and good government of theCommonwealth with respect to: “Copyrights, patents of inventionsand designs, and trade marks.”68The Varieties Act provided for the grant of plant varietyrights in new plant varieties for a period of twenty years.69 Therepeal of the Varieties Act by section 78 of the Breeder’s RightsAct also brought into operation transition provisions, particularlysection 82, which stated that plant variety rights under theprevious statute were to be treated as plant breeder’s rights underthe new statute. Pursuant to this statutory regime, the seconddefendant, Cultivaust Pty. Ltd., argued that:it was the licensee from the State of Tasmania of theFranklin barley rights;this gave it the exclusive right to sell and export Franklinbarley andthe plaintiff, in selling and exporting Franklin barleyfrom Australia, had acted in breach of the seconddefendant’s rights.70The plaintiff, Grain Pool of WA, defended against the claims byasserting that the rights asserted by the second defendant wereineffective because the grant of those rights was beyond theCommonwealth’s legislative powers in section 51 (xviii), (xxix) and(xxxix).71 The full bench of the High Court held that both Actswere within the powers contained in section 51 (xviii) of theConstitution and therefore it was unnecessary to consider whetherthe Acts could be supported under section 51 (xxix) or (xxxix).72The plaintiff had argued that the operation of section 51(xviii)68. Id. at para. 13.69. Plant Variety Rights Act, 1987, § 32 (Austl.) (commencing on the day ofacceptance of successful application).70. Pursuant to section 26 of the Plant Variety Rights Act, a grant was made onJanuary 19, 1990, in favour of the Tasmanian Department of Primary Industry andFisheries with respect to Franklin barley.71. Grain Pool, [2000] H.C.A. 14 at para. 12.72. Id. at para. 136.

with respect to “patents of inventions” was limited by certaintraditional principles of patent law. They argued that:there were certain fixed minimum requirements for the“intellectual effort” required of inventors respectingnovelty and inventive steps;there was a crucial distinction between process andproduct claims andthe term “patents” imports a constitutional requirementof the scope of the monopoly rights which must begranted and limits the permissible statutoryqualifications to those rights.73The court rejected those arguments, finding that the legislativescheme before it was sufficiently connected to the notion of“patents of invention.” While inventive step, novelty and exclusiverights might be determined differently than they would under thePatents Act, 1990, that did not mean that the legislation was notwith respect to the subject matter of “patents of invention.” Inother words, the fact that the legislation invoked unique notions ofnovelty, inventive step and exclusive rights in relation to plantbreeding did not mean the legislation was beyond parliamentarypower to legislate with respect to “patents of inventions.”At a broader level, the plaintiff contended that theenumerated head of legislative power in the constitution did notsupport the Varieties and Breeders Act.74 In particular, theyclaimed that the head of power should not be read so as to supportlegislation granting rights for products of every level ofintellectual effort.75The majority commenced its reasoning by quoting from thejoint judgment of the High Court in Nintendo Co. Ltd. v.Centronics Systems Pty. Ltd.,76 upholding the validity of theCircuit Layouts Act, 1989, where the court considered theconstruction of the terms of section 51(xviii):The grant of Commonwealth legislative power whichsustains the [Circuit Layouts Act] is that contained in s73. Id. at para. 12.74. Id. at para. 17.75. Id.76. (1994) 181 C.L.R. 134 (Austl.).

51(xviii) of the Constitution with respect to ‘Copyrights,patents of inventions and designs, and trade marks’. It isof the essence of that grant of legislative power that itauthorizes the making of laws which create, confer, andprovide for the enforcement of, intellectual propertyrights in original compositions, inventions, designs, trademarks and other products of intellectual effort.77The majority explained that an appropriate approach to theinterpretation of section 51(xviii) was that of Judge Higgins inAttorney-General for NSW v. Brewery Employes Union of NSW(“the Union Label Case”):78Higgins J observed in that case that trade marks were“artificial products of society” and that whilst “we are toascertain the meaning of ‘trade marks’ as in 1900,” trademarks usage in 1900 “gives us the central type; it doesnot give us the circumference of the power” with respectto trade marks provided for by s 51(xviii). The centre ofthe thing named—trade marks—was to be taken with themeaning as in 1900 to find the circumference of thepower. However, it would be “a mistake to treat thecentre as the radius.”79In line with Judge Higgins’s approach, the majority explainedthat the broad term “intellectual effort” embraced a variable,rather than a fixed, constitutional criterion and that the“origination” or “breeding” required respectively by the VarietiesAct and the Breeder’s Rights Act involved sufficient “intellectualeffort” to enliven the power.80Judge Kirby, in a separate judgement, advocated an approachdifferent from the majority regarding interpretation of the term“patents of inventions”:The power conferred by s 51(xviii) is a very broad one.The metes and bounds of the power are not to beascertained in my respectful view, by an understanding ofwhat fell within the class of “copyrights, patents ofinventions and designs, and trade marks” in 1900.77. Id. at 160; see Grain Pool, [2000] H.C.A. 14 at para. 17.78. (1908) 6 C.L.R. 469, 610-11 (Austl.).79. Grain Pool, [2000] H.C.A. 14 at para. 19.80. Id. at para. 42.

Because, in part of the joint reasons the other members ofthis Court have repeatedly referred to the considerationof the ambit of “patents of invention” in 1900—lest thatconsideration be thought to control, or even significantlyto influence, the contemporary meaning of theconstitutional words, I am bound to express my differentviewpoint . . . . [the] ultimate criterion was not pre-Federation legal understandings but a search for the“essential characteristics” of the words used in theConstitution . . . . The words used in the Constitution,and specifically those used in s 51(xviii), are not shackledto the legal understandings about the patents power in1900 . . . . Given the objects of the head of power, whichinclude the facilitation and protection of intellectualinventiveness within Australia, it would be speciallydestructive of the achievement of those objects if thegrant of power were to be attached—even as a primaryreference point—to the particular notions which, up to1900, “copyrights, patents of inventions and designs andtrade marks” had been protected by the law . . . . To thefull extent that the language of the Constitution warrantsand that other important values which it upholds permit,meaning should be given to a provision such as s 51(xviii)in a way that allows the section to respond to the verygreat variety of inventiveness that may be considered bythe Federal Parliament to necessitate protection for the“products of intellectual effort.”8181. Id. at paras. 90, 103, 132-33, 135 (emphasis added). Also consider thefollowing passage from Judge Kirby’s judgment:However, the advent of biogenetically engineered organisms and ofinventions in the field of information technology have stimulated anapparently increased willingness on the part of United States courts torecognise the way in which patents and analogous forms of legal protectioncan sometimes encourage technological innovation to the economic and socialbenefit of the United States and beyond. The specific inclusion of s 51(xviii) inthe Australian Constitution affords a further reason for assigning to s51(xviii) a meaning that permits the protection of “products of intellectualeffort” in the variety in which such products now manifest themselves andthe even greater variety in which they can be expected to appear in thefuture.Id. at para. 134.

C. The Length of Time a Property Right Will EndureThe United States Constitution contains the express wordsthat the IP rights legislated should be for “limited Times.” Itcould be read to be an explicit guarantee of the public domain. Inthe recent case of Eldred v. Reno,82 the constitutional validity ofthe Copyright Term Extension Act, 1998 (CTEA)83 legislationextending the U.S. copyright term from life of the author plus fiftyyears to life of the author plus seventy years,84 was questioned.The plaintiffs, who were in the business of publishing materialswhen copyright had expired, argued that the CTEA wasunconstitutional for three reasons:in both its prospective and retrospective applications theCTEA unjustifiably infringes upon freedom of speechprotected by the First Amendment;in its application to preexisting works, the CTEA violatesthe originality requirement of the Copyright Clause;in extending the term of subsisting copyrights, the CTEAviolates the “limited Times” requirement of the CopyrightClause.85The First Amendment argument, which was rejected, isconsidered below. On the issue of lack of originality, the courtdisposed of the issue saying:The plaintiff’s reason from this that the CTEA cannotextend an extant copyright because the copyrighted workalready exists and therefore lacks originality. Not so.Originality is what made the work copyrightable in thefirst place. A work with a subsisting copyright hasalready satisfied the requirement of originality and neednot do so anew for its copyright to persist. If theCongress could not extend a subsisting copyright for wantof originality, it is hard to see how it could provide for acopyright to be renewed at the expiration of its initialterm—a practice dating back to 1790 and not questioned82. 239 F.3d 372 (D.C. Cir. 2001).83. 17 U.S.C. § 302 (1994 & Supp. V 1999).84. This change was made to bring the U.S. copyright law into line with the EU.85. Eldred, 239 F.3d at 374.

even by the plaintiffs today.86The plaintiffs were also unsuccessful on the issue of “limitedTimes.” The court explained that, at widest, if “Congress were[sic] to make copyright protection permanent, then it surely wouldexceed the power conferred upon it by the Copyright Clause.”87 Itwent on to dispose of the argument at hand by saying:The present plaintiffs want a limit well short of the ruleagainst perpetuities, of course. And they claim to havefound it—or at least a bar to extending the life of asubsisting copyright—in the preamble of the CopyrightClause: “The Congress shall have power . . . To promotethe Progress of Science and useful Arts . . . .” Their ideais that the phrase “limited Times” should be interpretednot literally but rather as reaching only as far as isjustified by the preambular statement of purpose: If 50years are enough to “promote . . . Progress,” then a grantof 70 years is unconstitutional. Here the plaintiffs runsquarely up against our holding in Schnapper v. Foley, inwhich we rejected the argument “that the introductorylanguage of the Copyright Clause constitutes a limit oncongressional power.” The plaintiffs, however, disclaimany purpose to question the holding of Schnapper; indeed,they expressly acknowledge “that the preamble of theCopyright Clause is not a substantive limit on Congress’legislative power.” Their argument is simply that “theSupreme Court has interpreted the terms ‘Authors’ and‘Writings’ in light of that preamble, and that this Courtshould do the same with ‘limited Times.’”The problems with this argument are manifest. First,one cannot concede that the preamble “is not asubstantive limit” and yet maintain that it limits thepermissible duration of a copyright more strictly thandoes the textual requirement that it be for a “limitedTime.” Second, although the plaintiffs claim that Feistsupports using the preamble to interpret the rest of theClause, the Court in Feist never so much as mentions thepreamble, let alone suggests that the preamble informs86. Id. at 376-77.87. Id. at 377.

its interpretation of the substantive grant of power to theCongress (which there turned upon the meaning of“Authors” and of “Writings,” each standing alone).Similarly, the Trade-Mark Cases cited in Feist rest uponthe originality implied by “invention [and] discovery” andby the “writings of authors,” and make no reference at allto the preamble.88Such reasoning goes close to conflicting with the followingstatement of the Supreme Court in Graham v. John Deere Co.89“The Congress in the exercise of the patent power may notoverreach the restraints imposed by the stated constitutionalpurpose. Nor may it enlarge the patent monopoly without regardto the innovation, advancement or social benefit gainedthereby.”90 It also goes close to conflicting with the followingpassage from Morton Salt Co. v. G.S. Suppiger Co.:91The grant to the inventor of the special privilege of apatent monopoly carries out a public policy adopted bythe Constitution and laws of the United States, “topromote the Progress of Science and useful Arts, bysecuring for limited Times to . . . Inventors the exclusiveRight . . .” to their “new and useful” inventions. UnitedStates Constitution. But the public policy which includesinventions within the granted monopoly excludes from itall that is not embraced in the invention. It equallyforbids the use of the patent to secure an exclusive rightor limited monopoly not granted by the Patent Office andwhich it is contrary to public policy to grant.92In United States v. Moghadam,93 a case concerning theconstitutional validity of the anti-bootlegging statute (consideredin more depth below), the court made note of the apparentperpetual duration of the intellectual property rights grantedunder the statute:We note that there is another limitation in the CopyrightClause that may be implicated by the anti-bootlegging88. Id. at 377-78 (citation omitted).89. 383 U.S. 1 (1966).90. Id. at 5-6.91. 314 U.S. 488 (1942).92. Id. at 492 (citations omitted).93. 175 F.3d 1269 (11th Cir. 2000).

statute: the “Limited Times” requirement that forbidsCongress from conferring intellectual property rights ofperpetual duration. On its face, the protection created bythe anti-bootlegging statute is apparently perpetual andcontains no express time limit; therefore phonorecords oflive musical performances would presumably never fallinto the public domain. However, because Moghadamhas not challenged the constitutionality of § 2319A onthis basis, we decline to raise the issue sua sponte. Thus,we do not decide in this case whether extendingcopyright-like protection under the anti-bootleggingstatute might be fundamentally inconsistent with the“Limited Times” requirement of the Copyright Clause,and we do not decide in this case whether the CommerceClause can provide the source of Congressional power tosustain the application of the anti-bootlegging statute insome other case in which such an argument is preserved.We reserve those issues for another day. 94In Canada and Australia the issue of limited times is not assistedby express recognition of such a limit in the constitutionalmandate. However, upon further consideration of the matterthere could be little argument in Australia or Canada with theview that the granting of intellectual property rights is in someway limited by the reason for granting those rights.95 In fact, byway of tradition96 and under International law,97 there arelimits—statutory intellectual property rights (IPR) are generally94. Id. at 1281 (citations and footnote omitted).95. See Australian Trade Marks Act, 1995, §§17, 72, 75. Although some maysuggest statutory trademarks are of perpetual duration it is important to note they areusually only recognised where there is evidence of intended or actual use and areregistered for a period of ten years, which can be renewed.96. See Statute of Anne (1709) 8 Ann., c. 21 (Eng.); Statute of Monopolies, 1623,21 & 22 Jam. 1, c. 3 (Eng.); Millar v. Taylor, 98 Eng. Rep. 201, 231-32 (K.B. 1769).97. See Berne Convention for the Protection of Literary and Artistic Works (1971).International law can be used to assist in constitutional interpretation in Australia.See, e.g., Kartinyeri v. Commonwealth [1998] H.C.A. 22 (Austl.); “Applicant A” v.Minister for Immigration & Ethnic Affairs (1997) 142 A.L.R. 331 (Austl.); Kruger v.Commonwealth (1997) 146 A.L.R. 126 (Austl.); Newcrest Mining (WA) Ltd. v.Commonwealth (1997) 190 C.L.R. 513 (Austl.); Minister for Immigration & EthnicAffairs v. Teoh (1995) 183 C.L.R. 273 (Austl.); Horta v. Commonwealth (1994) 181 C.L.R.183 (Austl.); Bryan Horrigan & Brian F. Fitzgerald, International and TransnationalInfluences on Law and Policy Affecting Government, in Government Law and Policy § 1(Bryan Horrigan ed., 1998).

of a limited term and in certain circumstances subject to useswithout permission of the IPR owner—and based on the rationale(even if somewhat vague) that allowing monopoly rights ininformation benefits society as a whole through the creation ofvalue added informational products.98 Whether such a balance isachieved is a moot point.99 Vaver argues that “[i]f the rightsrestrict availability and use more than they increase them, theyare unjustifiable . . . .” 100In defining a patent, Judge Kirby in Grain Pool, Kirbyreferred to its limited nature. He explained:Upon this basis, the lawmaking power with respect to“patents of inventions” within s 51(xviii) involves theprovision by the state to the grantee of exclusive rightsfor a limited time to exploit, and to authorise otherpersons to exploit, a novel object or process of potentialbenefit to the community in respect of which a patent maybe granted and which is recorded in a public registerupon conditions of disclosure. This is the bedrock.Nothing more is required by the “really essentialcharacteristics” of “patents of inventions.”101While agreeing that the section 51 (xviii) head of powershould be read generously, Judge Kirby made it clear thatintellectual property rights should not be regarded as absolute;they must be constructed in the context of their proprietarypurpose. He explained “the protection of intellectual propertyrights must be afforded in a constitutional setting which upholdsother values of public good in a representative democracy.”102In contrast, by suggesting Article 1, Section 8, Clause 8 of theUnited States Constitution is narrower than section 51 (xviii) inthat it requires a purpose of promoting the progress of science andthe useful arts, the majority in Grain Pool appears to say that avariety of intellectual property rights (from low to highintellectual input) can be legislated under the section 51 (xviii)98. See Landes, supra note 12 (discussing the utilitarian/economic nature of suchthinking); see also Sony, 464 U.S. at 429-32; Vaver, supra note 19, at 87-96 (noting therelevance of moral rights in Canadian copyright law).99. See Vaver, supra note 19, at 6-8.100. Id. at 8.101. Grain Pool, [2000] H.C.A. 14 at para. 135 (emphasis added).102. Id. at n.218.

head of power. It is not clear whether the majority is alsosuggesting that these rights, once legislated, have anyconstitutional limit in the name of the public domain.103Likewise, the following passage from Judge Higgins in the UnionLabel case might suggest that there are no constitutional limits onintellectual property rights: “The power to make laws ‘withrespect to’ these rights involves the power to declare what shall bethe subject of the rights.”104 But that is far from clear.Further conjecture on this issue was raised by the High Courtin explaining the notion of “property” in Yanner v. Eaton:105It refers to a degree of power that is recognised in law aspower permissibly exercised over the thing. The conceptof “property” may be elusive. Usually it is treated as a“bundle of rights.” But even this may have its limits asan analytical tool or accurate description, and it may be,as Professor Gray has said, that “the ultimate fact aboutproperty is that it does not really exist: it is mereillusion.” Considering whether, or to what extent, therecan be property in knowledge or information or propertyin human tissue may illustrate some of the difficulties indeciding what is meant by “property” in a subjectmatter.106There is strong indication in this passage that informationalrights do have boundaries. Furthermore, the passage expresslyacknowledges that power is inherent in the notion of property. Asit is the function of law to mediate power relations in the name ofa prosperous, humane and ordered society, one could logicallyassert that property will have some legally defined limits.Lessig argues that just as real property has limits, e.g., it canbe zoned, so does intellectual property, e.g., fair use in copyrightlaw and these limits, when imposed, ensure that the collectiveinterest of society to access or use are privileged over the absolute103. Id. at para. 32 (“In particular it will be open to the Parliament to pursue itspolicies by legislation with respect to various subject-matters, if one of them appears ins 51 (xviii) on an adequate reading of that text.”). Id.104. (1908) 6 C.L.R. 469 at 611 (Austl.).105. (1999) 201 C.L.R. 351 (Austl.).106. Id. at 366; see generally Michael A. Heller, The Boundaries of Private Property,108 Yale L.J. 1163 (1999) (arguing that private property is not a useful analyticconcept unless it can be practically bounded).

interest of the property holder.107 Likewise, Judge Kirby’sapproach challenges us to conceptualise a more principledconstruct of intellectual property in accordance with the notionthat “no right is absolute.”108 Are we to expect from suchreasoning a view that demands (in a constitutional sense) thatintellectual property rights must yield to the public domain wherenecessary? And will such reasoning support the notion that asintellectual property rights have had and still do have limitedterms (based on a mixture of economic utility and the democraticand social value of the public domain),109 the constitutional powerto create such rights is subject to limits, specifically, arequirement that intellectual property rights be granted forlimited terms?110 Further, will we see arguments suggesting thatthere is a constitutional guarantee to the public domain?111The Grain Pool case raises interesting questions concerningintellectual property and constitutional principle for theinformation age. It shows the willingness of the High Court ofAustralia to allow the Commonwealth or Federal Parliament tolegislate on issues of intellectual product and gives insight intothe variable definition of the term “patents of invention.” On thisplane, the High Court is clearly saying that, in terms of thefederal division of legislative powers, the CommonwealthParliament has a broad power to legislate on matters relating tointellectual effort. On another plane, Judge Kirby’s judgement isextremely interesting in that it suggests that intellectual propertyrights do have constitutionally defined limits. In other words, the107. See generally Lessig, supra note 42, at 131-35 (1999) (discussing the limits onthe protection of property).108. See generally Schenck v. United States, 249 U.S. 47, 52 (1919) (quotingHolmes, J., “The most stringent protection of free speech would not protect a man infalsely shouting fire in a theatre and causing a panic.”).109. Brian F. Fitzgerald, Underlying Rationales of Fair Use: Simplifying theCopyright Act, 2 S. Cross U. L. Rev. 153 (1998), available athttp://www.scu.edu.au/schools/lawj (last visited Nov. 3, 2001).110. See, e.g., Lessig, supra note 42, at 133-34 (explaining that in the UnitedStates, exclusive copyrights are granted for a limited term in accordance with theUnited States Constitution, and arguing the term should be limited by and linked tothe purpose for which the exclusive rights are granted, progressing not hindering theuseful arts); see generally Eldred, 239 F.3d at 372 (challenging the extension of U.S.copyright term to life of the author plus seventy years).111. See generally Lessig, supra note 42, at 133-34 (arguing that the public domainis implied in the United States Constitution as a consequence of the requirement thatexclusive rights be for limited times).

elationship of power between the Commonwealth Parliament andthe people in legislating intellectual property rights is notabsolute.112 In their silence on this issue, the majority raisesconcerns and uncertainty as to the limits of intellectual propertyrights in the information age.113 We must be mindful that in anera when information is a core economic and cultural resource, itis vital as democratic principle that the legal process ofcommodifying information through intellectual property law beclearly defined and understood.D. Fixed in a Tangible Medium?A hallmark of copyright and patent law is that the intellectualknowledge must be embodied, manifested or fixed in a tangiblemedium. A recent Australian report by the Copyright LawReform Commission suggests that in the realm of copyright lawthe whole notion of fixation in a tangible medium is redundant inthe digital age and should be removed or at least reconceptualisedas a criterion for copyright protection.114While there is no express requirement for fixation in atangible medium in the Australian or Canadian Constitutions, theword “writings” in the United States clause implies some notion offixation. The issue was raised but not finally resolved inMoghadam. That case concerned the constitutional validity of ananti-bootlegging statute115 criminalizing the unauthorizedrecording, transmission to the public, and sale or distribution of or112. See generally Brian F. Fitzgerald, Proportionality and AustralianConstitutionalism, 12 U. Tas. L. Rev. 263 (1993) (discussing the notion of sovereignty inAustralian constitutional law).113. See, e.g., Carl Shapiro & Hal R. Varian, Information Rules: A Strategic Guideto the Network Economy (1999); Benkler, supra note 10; Felix S. Cohen,Transcendental Nonsense and the Functional Approach, 35 Colum. L. Rev. 809, 814-17(1935); Rochelle C. Dreyfuss, We Are Symbols and Inhabit Symbols, So Should We BePaying Rent? Deconstructing the Lanham Act and Rights of Publicity, 20 Colum.-VLAJ. L. & Arts 123 (1996); Rochelle C. Dreyfuss, Expressive Genericity: Trademarks asLanguage in the Pepsi Generation, 65 Notre Dame L. Rev. 397 (1989); Brian F.Fitzgerald, Conceptualising the Digital Environment, in Going Digital 2000: LegalIssues for E-Commerce Software and the Internet 1 (Anne Fitzgerald et al. eds., 2d ed.2000); Michael A. Heller, The Tragedy of the Anticommons: Property in the TransitionFrom Marx to Markets, 111 Harv. L. Rev. 621 (1998); David Lange, Recognising ThePublic Domain, 44 Law & Contemp. Probs. 147 (1981); Ejan Mackaay, The Economicsof Emergent Property Rights on the Internet, in The Future of Copyright in a DigitalEnvironment 13 (P. Bernt Hugenholtz ed. 1996).114. Copyright Law Review Committee, supra note 60, at paras. 5.48-.53.115. 18 U.S.C. § 2319A (1994 & Supp. V 1999).

traffic in unauthorized recordings of live musical performances.116Moghadam was convicted of violating the statute after he pleadedguilty to knowingly distributing, selling, and trafficking in bootleg(unauthorized) compacts discs featuring live musicalperformances by recording artists including Tori Amos and theBeastie Boys. Moghadam then sought to challenge theconstitutional validity of the statute arguing it did not fall withinany of the federal legislative powers enumerated in Article I,Section 8, Clause 8 of the Constitution.117 The governmentresponded that it was constitutional under either the CopyrightClause or the Commerce Clause.118 The anti-bootlegging statute,which was enacted as a consequence of and in implementation ofthe TRIPS agreement provides in 18 U.S.C. section 2319A that:(a) Whoever, without the consent of the performer orperformers involved, knowingly and for purposes ofcommercial advantage or private financial gain–(1)fixes the sound or sounds and images of a live musicalperformance in a copy or phonorecord, or reproducescopies or phonorecords of such a performance from anunauthorized fixation;(2)transmits or otherwise communicates to the public thesounds or sounds and images of a live musicalperformance; or(3)distributes or offers to distribute, sells or offers to sell,rents or offers to rent, or traffics in any copy orphonorecord fixed as described in paragraph (1),regardless of whether the fixations occurred in the UnitedStates;shall be imprisoned . . . or fined . . . or both . . . .119The Act also contains a similar provision establishing civil liabilityfor the same conduct but omitting the commercial advantage orprivate financial gain requirement.120 There is little legislative116. Moghadam,175 F.3d at 1271.117. Id.118. Id.119. Id. at 1272.120. See 17 U.S.C. § 1101 (1994).

history dealing with either provision because the Act was rushedthrough Congress on fast-track procedures, yet what existssuggests that Congress viewed the anti-bootlegging provisions asenacted pursuant to its Copyright Clause authority.121The court explained that the rights created by the antibootleggingprovisions were actually hybrid rights that in someways resembled the protection of copyright law but in other wayswere distinct.122 The court explained that it was unclear whetherlongstanding concepts generally applicable to copyright law suchas fair use, the work-for-hire doctrine and limited duration, applyto the anti-bootlegging provisions.123 The court noted that incontrast to the six exclusive rights of a copyright owner spelled outin 17 U.S.C. section 106, it appeared that the only exclusive rightcreated by the anti-bootlegging statute is to record and/or recommunicateone’s performance.124 For all of those reasons, thecourt concluded that the protection the anti-bootlegging statuteconfers on musicians is best described as “quasi-copyright” or suigeneris protection.125Judge Anderson, writing the opinion for the court, held thestatute to be validly enacted under the Commerce Clause. On theissue of the Copyright Clause, the court explained:This positive grant of legislative authority includesseveral limitations. Of these limitations, Moghadam hasrelied in the instant case only on the concept of “fixation”which is said to be embedded in the term “Writings.”The concept of fixation suggests that works are notcopyrightable unless reduced to some tangible form. “Ifthe word ‘writings’ is to be given any meaningwhatsoever, it must, at the very least, denote somematerial form, capable of identification and having amore or less permanent endurance.” Of course, the term“Writings” has been interpreted so broadly as to includemuch more than writings in the literal sense, or the lay121. 140th Cong. Rec. H11441, H11457 (daily ed. Nov. 29, 1994) (statement of Rep.Hughes) (“There are a number of changes in copyright that will advance our interestsin the area of bootlegging, which is going to basically protect our country.”).122. Moghadam, 175 F.3d at 1272.123. Id. at 1273.124. Id.125. Id.

definition of the word. In fact, since a sound recordingqualifies as a “Writing” in the constitutional sense, “it isnow clear that a writing may be perceptible eithervisually or aurally . . . .” But the fixation requirementseems to have persisted through this expansion. Thus,although in the modern era the term “Writings” allowsCongress to extend copyright protection to a great manythings, those things have always involved some fixed,tangible and durable form.Moghadam argues that a live performance, by definition,has not been reduced to a tangible form or fixed as of thetime of the performance. Moghadam argues that, but forthe bootlegger’s decision to record, a live performance isfleeting and evanescent.Because we affirm the conviction in the instant case onthe basis of an alternative source of Congressional power,we decline to decide in this case whether the fixationconcept of Copyright Clause can be expanded so as toencompass live performances that are merely capable ofbeing reduced to tangible form, but have not been. Forpurposes of this case, we assume arguendo, withoutdeciding, that the above described problems with thefixation requirement would preclude the use of theCopyright Clause as a source of Congressional power forthe anti-bootlegging statute.126E. The Commerce ClauseThe enumerated powers in the United States and AustralianConstitutions are designed to allocate legislative responsibilitiesbetween the federal and state governments. They do notnecessarily limit governmental power or bestow rights onindividuals, although clearly one provision in the AustralianConstitution does—section 51 (31). In recent times seriousarguments have been made—in the context of proposed databaselegislation and the Digital Millennium Copyright Act(DMCA)127—that the specific requirements of the IP clause define126. 175 F.3d at 1273-74.127. 17 U.S.C. §§ 1201-1332 (Supp. IV 1998).

the parameters of the propertization of information throughlegislative means.128 In essence, the argument is that the IPclause is both a grant of power and a limitation on governmentpower ensuring such fundamental information rights of user andaccess, such as the right to ideas, the public domain and fairuse.129An even more interesting argument arising from thisapproach is whether state governments are in any way limited bythe specific requirements of the IP clause. In other words, doesthe IP clause in any way preempt state legislation on this topic?Goldstein held that the American states have, subject topreemption by legislation,130 a concurrent power to legislate uponintellectual property rights.131 In that case, the Supreme Courtimplied that the states are not constrained by any limits in thefederal IP clause as the legislation in question, which the Courtupheld, allowed perpetual duration of copyright—in contrast tothe limited times requirement in the federal clause.132The clause most likely to undermine the IP clause at thefederal level is the Commerce Clause. The United StatesCommerce Clause reads that Congress has the legislativeauthority “[t]o regulate Commerce with foreign Nations, andamong the several States . . . .”133 The Australian CommerceClause reads that the Commonwealth Parliament may make lawsfor the peace, order and good government of the Commonwealthwith respect to trade and commerce with other countries andamong the States.134In Moghadam, the court refused to decide whether the antibootleggingstatute could be upheld under the Copyright Clause;they were of the view that it came within the Commerce128. See Marci A. Hamilton, Database Protection and the Circuitous Route Aroundthe United States Constitution, in International Intellectual Property and the CommonLaw World (Charles E.F. Rickett & Graeme W. Austin eds., 2000); see also, Benkler,supra note 10, at 440-46; Paul J. Heald & Suzanna Sherry, Implied ConstitutionalLimits on Congressional Power: Construing the Commerce Power in Light of theIntellectual Property Clause, 2000 U. Ill. L. Rev. 1119 (2000); William Patry, TheEnumerated Powers Doctrine and Intellectual Property: An Imminent ConstitutionalCollision, 67 Geo. Wash. L. Rev. 359 (1999).129. See John Deere Co., 383 U.S. at 5.130. See, e.g., 17 U.S.C. § 301 (1994 & Supp. V 1999).131. Goldstein, 412 U.S. at 560.132. Id. at 560-61.133. U.S. Const. art I, § 8, cl. 3.134. Aust. Const. § 51(i).

Clause.135 The court reviewed the two opposing precedents inlight of the Commerce Clause in the face of other clauses andconcluded that it was not against the spirit of the Constitution toallow the statutes to be upheld under the Commerce Clause. Theyreasoned that this type of legislation supplemented copyrightprotection and should be allowed.136The court started out by confirming that section 2319Aconcerned conduct that had a substantial effect on commercebetween the several states and commerce with foreign nations,namely the sale of bootleg compact discs. The court explained thatthe more interesting question was whether Congress could use theCommerce Clause to avoid the limitations that might prevent itfrom passing the same legislation under the Copyright Clause.They explained:that in general, the various grants of legislative authoritycontained in the Constitution stand alone and must beindependently analyzed. In other words, each of thepowers of Congress is alternative to all of the otherpowers, and what cannot be done under one of them mayvery well be doable under another. 137The court noted that this general approach had been appliedpreviously in a context involving the Copyright Clause and theCommerce Clause, namely, the Trade-Mark Cases.138 In thosecases, the constitutional validity of an early trademark lawenacted by Congress in 1876 was at issue. The Supreme Courtheld that the Copyright Clause could not sustain the 1876 Actbecause “[t]he ordinary trade-mark has no necessary relation toinvention or discovery,” which were the hallmarks of protectablesubject matter under the Copyright Clause.139 The Courtexplained that trademarks were inherently commercial; theconcept behind the 1876 Act (and modern trademark law) was notto encourage intellectual and artistic development, but rather toprotect businesses from predatory behaviour in themarketplace.140 This concept was said to be outside the ambit of135. Moghadam, 175 F.3d at 1269.136. Id.137. Id. at 1277.138. See United States v. Steffens, 100 U.S. 82 (1879).139. Id. at 94.140. See id. at 94-99.

the legislative power contained in the IP clause.141However, that was not the end of the matter and the SupremeCourt proceeded to test the validity of the statute against theCommerce Clause. As the Commerce Clause at that time wasnarrowly construed, the legislation was not upheld. Since thattime, an expanding view of the ambit of the Commerce Clausemeans that today the modern variant of this trademark law, theLanham Act, is seen to be an exercise of the commerce clause.142The Eleventh Circuit Court of Appeals summarised this point,saying that “although the 1876 Act did not survive due to therestrictive view of the Commerce Clause prevailing at that time,the Supreme Court’s analysis in the Trade-Mark Cases stands forthe proposition that legislation, which would not be permittedunder the Copyright Clause, could nonetheless be permitted underthe Commerce Clause, provided that the independentrequirements of the latter are met.”143However, the Court noted that another line of authority stoodin contrast to the Trade-Mark Cases and needed moreconsideration. In Railway Labor Executives’ Ass’n v. Gibbons,144the Supreme Court considered a statute enacted by Congress thatpurported to alter a pending bankruptcy case by requiring thedebtor railroad company’s bankruptcy estate to pay $75 million tothe company’s former employees. That statute directly clashedwith the Bankruptcy Clause of the United States Constitution,145which provides that Congress is empowered to pass “uniform”bankruptcy laws, because the law targeted a particular situationand was anything but uniform. The Supreme Court rejected theview that the legislation could nevertheless be sustained underthe Commerce Clause, which contains no uniformity requirement,stating that “if we were to hold that Congress had the power toenact nonuniform bankruptcy laws pursuant to the CommerceClause, we would eradicate from the Constitution a limitation onthe power of Congress to enact bankruptcy laws.”146 The Court141. See id. at 96-98.142. See Michael B. Gerdes, Comment, Getting Beyond Constitutionally MandatedOriginality as a Prerequisite for Federal Copyright Protection, 24 Ariz. St. L.J. 1461,1467 (1992).143. Moghadam, 175 F.3d at 1278.144. 455 U.S. 457 (1982).145. U.S. Const. art. I, § 8, cl. 4146. Gibbons, 455 U.S. at 468-69.

econciled the authorities, saying:We note that there is some tension between the formerline of cases (Heart of Atlanta, the Trade-Mark Cases andAuthors League) and the Railway Labor Executives case.The former cases suggest that in some circumstances theCommerce Clause can be used by Congress to accomplishsomething that the Copyright Clause might not allow.But the Railway Labor Executives case suggests that insome circumstances the Commerce Clause cannot be usedto eradicate a limitation placed upon Congressional powerin another grant of power. For purposes of the instantcase, we resolve this tension in the following manner. Inresolving this tension and in reaching our conclusion inthis case, we undertake a circumscribed analysis,deciding only what is necessary to decide this case, andwe reach a narrow conclusion. First, as described above,we hold the anti-bootlegging statute satisfies the“substantial effects” test of the post-Lopez CommerceClause jurisprudence. Second, following the former lineof cases (Heart of Atlanta, the Trade-Mark Cases andAuthors League), we hold that in some circumstances theCommerce Clause indeed may be used to accomplish thatwhich may not have been permissible under theCopyright Clause. We hold that the instant case is onesuch circumstance in which the Commerce Clause may bethus used. It is at this point that we must resolve thetension with Railway Labor Executives.Resolving this tension, we take as a given that there aresome circumstances, as illustrated by Railway LaborExecutives, in which the Commerce Clause cannot beused by Congress to eradicate a limitation placed uponCongress in another grant of power. For the reasons thatfollow, we hold that the instant case is not one suchcircumstance. We hold that the Copyright Clause doesnot envision that Congress is positively forbidden fromextending copyright-like protection under otherconstitutional clauses, such as the Commerce Clause, toworks of authorship that may not meet the fixationrequirement inherent in the term “Writings.” The grantitself is stated in positive terms, and does not imply any

negative pregnant that suggests that the term “Writings”operates as a ceiling on Congress’ ability to legislatepursuant to other grants. Extending quasi-copyrightprotection to unfixed live musical performances is in noway inconsistent with the Copyright Clause, even if thatClause itself does not directly authorize such protection.Quite the contrary, extending such protection actuallycomplements and is in harmony with the existing schemethat Congress has set up under the Copyright Clause. Alive musical performance clearly satisfies the originalityrequirement. Extending quasi-copyright protection alsofurthers the purpose of the Copyright Clause to promotethe progress of the useful arts by securing some exclusiverights to the creative author. Finally, with respect to thefixation requirement, upon which this opinion focuses,although a live musical performance may not have beenfixed, or reduced to tangible form, as of the time thebootleg copy was made, it certainly was subject to havingbeen thus fixed. Our conclusion that extending copyrightlikeprotection in the instant case is not fundamentallyinconsistent with the fixation requirement of theCopyright Clause is bolstered by an example from theprior copyright law. If a live performance is broadcast,e.g., by radio or television, and simultaneously recordedby the performer, any unauthorized recording by a personreceiving the broadcast constitutes copyrightinfringement of the sound recording or motion picture,notwithstanding that the infringer actually copied thelive performance directly, and not the fixation thereof.This result is based upon the last sentence of thedefinition of “fixed” in 17 U.S.C. § 101. That last sentenceprovides: “A work consisting of sounds, images, or both,that are being transmitted, is ‘fixed’ for purposes of thistitle if a fixation of the work is being madesimultaneously with its transmission.” This definitioncreates a legal fiction that the simultaneous fixationoccurs before the transmission and the unauthorizedrecording . . . . While we are aware that theconstitutionality of this aspect of the statute has neverbeen tested, the ease with which it has been incorporatedinto the prior copyright law suggests that fixation, as a

constitutional concept, is something less than a rigid,inflexible barrier to Congressional power. Indeed, if aperformer under the prior law could effectively protect alive musical performance, circumventing the fixationrequirement, simply by the device of simultaneousrecordation, the anti-bootlegging law seems to us likemore of an incremental change than a constitutionalbreakthrough. Common sense does not indicate thatextending copyright-like protection to a live performanceis fundamentally inconsistent with the Copyright Clause.For the foregoing reasons, we conclude that extendingcopyright-like protection in the instant case is notfundamentally inconsistent with the fixation requirementof the Copyright Clause. By contrast, the nonuniformbankruptcy statute at issue in Railway Labor Executiveswas irreconcilably inconsistent with the uniformityrequirement of the Bankruptcy Clause of theConstitution. 147The question remains as to whether a sui generis database lawenacted pursuant to the Commerce Clause would surviveconstitutional challenge. The argument would be made that theIP clause requires originality (a modicum of creativity andsomething that emanates from the author) and that allowinginformation to be propertized in situations where originality is notevident would be unconstitutional.148 The focus of the argumentis whether the IP clause sets the parameters for propertizinginformation through its specific requirements. One problem withthis approach is that the IP clause was established in another era.Nowadays, information is a lifeblood, both in social and economicterms, and it is conceivable that new ways of propertizinginformation have emerged and will continue to emerge that willdemand protection; yet the IP clause may not be up to thischallenge in conceptualisation. A consequence of limiting theCommerce Clause in this respect may be that interested partieswill resort to the states to enact their legislation (difficult, in theface of section 301 of the Copyright Act). This may result in lessthan desirable outcomes. While it is clear that restricting the147. Moghadam, 175 F.3d at 1279-81.148. See supra note 128.

scope of the Commerce Clause in the face of the IP clause hasmerit and legal precedent to support it, an over-zealousimplementation of such a principle could mortify constitutionalinterpretation, stifle incentive to create and ensuing innovationand retard thinking about new forms of digital property.In Australia, a similar question concerning the commerceclause and IP clause could arise although the specificrequirements of the U.S. clause are not explicit and possibly notimplicit in the Australian clause. This is nevertheless a dormantissue that needs to be considered when legislating in this area.149In summary, note that specific limits inherent in the IP clausemay be asserted in constitutional litigation. The law in light ofMoghadam appears to suggest that specific limits pertaining tothe IP clause will limit the Commerce Clause in certain cases andthat this might be an obstacle for any sui generis database law inthe United States.II. THE CONSTITUTIONAL IMPERATIVE: FREE SPEECH, ACCESS TOINFORMATION AND DEMOCRATIC PRINCIPLEThe second dimension of traditional constitutional principlethat promises to define, if not influence, the boundary of digitalproperty is that of free speech: constitutionally and expresslyenshrined in the United States First Amendment, and Article 2 ofthe Canadian Charter of Rights150 and, existing to a limited andimplied degree, in the context of political speech in Australia.151As to this point, the First Amendment to the United StatesConstitution, which has been vigorously asserted in recent timesby advocates fearful of growing intellectual property rights, hasnot been considered by the courts to be of overriding concern.This is not to say that free speech considerations have notinfluenced the development and implementation of intellectualproperty law. The notion of free speech is implicated in thedefinition or construction of digital property because we usecopyrighted/patented/trademarked items in the process of149. See Australian Tape Mfrs. v. Commonwealth of Australia (1993) 176 C.L.R. 480(Austl.) (while the general rule is that each head of legislative power stands alone, theHigh Court has interpreted section 51 (xxxi) of the Australian Constitution (theAustralian version of the U.S. Takings Clause) as both a grant of power and alimitation on other powers in certain circumstances).150. See Fewer, supra note 19, at 226-39.151. See Lange v. Australian Broadcasting Corp. (1997) 145 A.L.R. 96 (Austl.).

communicating—in constructing knowledge and meaning. Themost obvious example might be when I copy an image and displayit on my website. If that use is purely commercial then anychance of a fair dealing or fair use defence is weak, yet the moreits shades into political or social commentary or research the morelikely we are to have an allowable use.Software which is integral to the digital communicativeprocess is also an interesting example. Software is a form ofdiscourse, making things manifest or allowing them to be seen; itis part of the speech or communicative process in the digitalenvironment. If I propertize software through copyright or patent,I allow a certain degree of control over communication.In the United States, those adamant that IP rights mustaccommodate the necessity of consuming information anddemocratically distributed intelligence—the cultural or socialaspects as opposed to the economics of information—have focussedon First Amendment arguments. In order to speak andcommunicate or think and construct knowledge in the new digitalenvironment, we need to be able to access and distributeinformation. “A broad dissemination of principles, ideas, andfactual information is crucial to the robust public debate andinformed citizenry that are ‘the essence of self-government.’ Andevery citizen must be permitted freely to marshal ideas and factsin the advocacy of particular political choices.”152 This is theessence of the debate, although the desire for informational libertyor diversity is not limited to just the political realm but alsoencompasses the economic, social and cultural plane.A. Reconciling Free Speech and CopyrightIn Harper & Row Inc. v. Nation Enterprises, the UnitedStates Supreme Court explained “[t]he Second Circuit noted,correctly, that copyright’s idea/expression dichotomy “[strikes] adefinitional balance between the First Amendment and theCopyright Act by permitting free communication of facts whilestill protecting an author’s expression.” No author may copyrighthis ideas or the facts he narrates.”153 In another passage theCourt referred to “copyright as the engine of free expression,” as it152. Harper & Row Publishers Inc. v. Nation Enters., 471 U.S. 539, 582 (1985)(Brennan, J., dissenting ) (quoting Garrison v. Louisiana, 379 U.S. 64, 74-75 (1964)).153. Harper & Row, 471 U.S. at 556.

gave incentive for people to express ideas.154 The existence of thefair use, along with other exceptions, was also important.155The defendants in Harper & Row had extracted/copied 300words from former President Gerald Ford’s biography, before itwas published, ruining a deal that had been made with TimeMagazine to do a pre-publication piece of 12,000 words. TheSupreme Court held that copying 300 words was not a fair use asit contained the most sensational details, including informationabout former President Nixon’s pardon (quality not quantitymattered here), and ruined an aspect of the commercialexploitation of the work. The defendants argued for a specialpublic figure or interest exception on top of fair use that wouldallow copyrighted work to be used in social communication. TheCourt rejected that argument, noting that copyright was itself anengine of free expression making it possible and rewarding forpeople to invest in the development and dissemination of ideas.156As the Act stood, there was sufficient reconciliation of thecommodifying or propertizing of information through copyrightand the notion of free speech.The First Amendment challenge to intellectual propertylegislation was dealt a further blow in the Eldred decision whenthe United States Court of Appeals for the District of Columbiaendorsed and reaffirmed the holding in Harper & Row that theCopyright Act and the First Amendment are adequately reconciledthrough the notions of the idea/expression dichotomy and the fairuse (along with other permissive use) doctrine(s).157 In otherwords, since the Copyright Act protects expression of ideas andnot the ideas themselves, there is technically no barrier to anycommunicative activity. Furthermore, fair use makes sociallynecessary material available in the face of the exclusive rights ofthe owner.In Eldred, the plaintiffs argued that the CTEA violated theFirst Amendment and was unconstitutional.158 The Courtexplained:The decisions of the Supreme Court in Harper & Row154. Id. at 558.155. Id. at 560.156. Id. at 558.157. Eldred, 239 F.3d at 375.158. Id. at 374.

Publishers Inc. v. Nation Enterprises and of this court inUnited Video, Inc. v. FCC, stand as insuperable bars toplaintiffs’ first amendment theory. In Harper & Row theCourt held that a magazine’s advance publication ofexcerpts from the memoirs of former President GeraldFord infringed the copyright thereon. In doing so theCourt explained how the regime of copyright itselfrespects and adequately safeguards the freedom of speechprotected by the First Amendment . . . . The [F]irst[A]mendment objection of the magazine was misplaced“[i]n view of the First Amendment protections alreadyembodied in the Copyright Act’s distinction betweencopyrightable expression and uncopyrightable facts andideas, and the latitude for scholarship and commenttraditionally afforded by fair use.” In keeping with thisapproach, we held in United Video that copyrights arecategorically immune from challenges under the FirstAmendment. There, certain cable companies petitionedfor review of an FCC regulation providing that thesupplier of a syndicated television program could agree tothe program being broadcast exclusively by a singlestation in a local broadcast area. We rejected the firstamendment aspect of their challenge as follows:In the present case, the petitioners desire to makecommercial use of the copyrighted works of others. Thereis no first amendment right to do so. Although there issome tension between the Constitution’s [C]opyright[C]lause and the first amendment, the familiaridea/expression dichotomy of copyright law, under whichideas are free but their particular expression can becopyrighted, has always been held to give adequateprotection to free expression.159The plaintiffs argued that those authorities were restrictedsolely to the narrow case where a litigant demands a right to useotherwise legitimately copyrighted material, which was distinctfrom the litigation at hand that sought to challenge theconstitutional validity of the Copyright Act itself. The Court159. Id. at 375-76 (citations omitted).

ejected this purported distinction as being “totally illusory.”160The Court explained, “the relevant question under the FirstAmendment—regardless whether it arises as a defence in a suitfor copyright infringement or in an anticipatory challenge to astatute or regulation—is whether the party has a FirstAmendment interest in a copyrighted work.”161The type of argument or justification for reconciling copyrightand the First Amendment given in Harper & Row and Eldred ismore difficult to sustain when a law proposes to protect facts oreliminates fair use. The European Database Directive of 1996protects facts and gives very limited fair use rights, as do proposeddatabase laws currently before the U.S. Congress. How mightthey stand up in the face of a First Amendment challenge? Thescrutiny or balancing process here will be interesting.Furthermore, Universal City Studios v. Reimerdes,162 concerningthe operation of the DMCA, suggests that the DMCA can act toprotect facts and eliminate fair use rights where a technologicalprotection measure is in place.163 The DMCA does this byprohibiting actual circumvention of an access control and byprohibiting providing, offering or otherwise trafficking in a devicethat will circumvent an effective technological access or copyprotection measure.164 It is no defence to this law to argue thatthe circumvention has taken place in order to make fair use ofmaterial behind the technological protection measure or to gainaccess to raw data or facts behind the technological protectionmeasure.165In Reimerdes, the issue of free speech was raised in thecontext of the DMCA anti-circumvention provisions. The ContentScrambling System (CSS) was a form of encryption used to protectDVDs from being played on unauthorised players. An authorisedDVD player, which included a computer running MicrosoftWindows, would allow the DVD to be played but not copied. Inorder to allow people to speak in different digital voices, namelythe Linux open code operating system, DeCSS was released.Initially, the 2600 website made the program available from their160. Id. at 376.161. Id.162. 111 F. Supp. 2d 294 (S.D.N.Y 2000).163. Id. at 322.164. See 17 U.S.C. §§ 1201(a)(1), (a)(2), (b) (Supp. IV 1998).165. Reimerdes, 111 F. Supp. 2d at 321-23; see Benkler, supra note 10, at 412-46.

site, but after an interim injunction was issued they removed itfrom their site and linked to other sites. Judge Kaplan held thatthey breached the DMCA by providing or otherwise trafficking ina device that circumvents a technological protection measure.166One series of arguments in the case, which is now on appealto the second circuit, was that the DMCA was unconstitutional onFirst Amendment grounds, as it restricted the expression of thecomputer code. Judge Kaplan held that while code in object orsource code can be speech for First Amendment purposes, thatdoes not mean it cannot be regulated.167 He explained that whilethe code is expressive, the behaviour it causes is functional andthe functional aspect of speech can be regulated even if it has anincidental impact on the expressive aspect.168 He explained thatregulation of the functional aspect of speech is content neutral, notcontent specific, and therefore only requires an intermediate, asopposed to a strict, level of scrutiny.169 The government objectivehere was to prevent large scale copyright piracy in the digitalenvironment—this was a legitimate objective and the scrutiny wassatisfied.I suggest the judge understated the discursive aspect of thefunctional behaviour of code. Code expressed as 1s and 0s isexpressive, it is a literary text for copyright purposes—that muchis obvious. But the behaviour that the code causes is not justfunctional, it is also communicative and discursive. Thebehaviour created by this code allowed the people using DeCSS to166. Reimerdes, 111 F. Supp. 2d at 319.167. Id. at 326.Defendants’ assertion that computer code is “protected” by the FirstAmendment is quite understandable. Courts often have spoken of certaincategories of expression as “not within the area of constitutionally protectedspeech,” so defendants naturally wish to avoid exclusion by an unfavorablecategorization of computer code. But such judicial statements in fact are notliterally true. All modes of expression are covered by the First Amendmentin the sense that the constitutionality of their “regulation must bedetermined by reference to First Amendment doctrine and analysis.”Regulation of different categories of expression, however, is subject to varyinglevels of judicial scrutiny. Thus, to say that a particular form of expression is“protected” by the First Amendment means that the constitutionality of anyregulation of it must be measured by reference to the First Amendment. Insome circumstances, however, the phrase connotes also that the standard formeasurement is the most exacting level available.Id. (footnotes and citations omitted).168. Id. at 327.169. Id. at 328.

talk, communicate or consume information in differently codedspeech, namely Linux. It may be that this communicative activitycan be regulated because it could lead to unlawful activity, butthat is a question for further consideration and it may be morecrucial in other cases. We need to keep in mind that software isnot just code, it is discourse; it is not just a noun, it is also a verb.Software allows us to see and say things in digital space. Itsexpressive or discursive potential is not just in code written asstatic text but in the running code, allowing and facilitatingpeople to move and consume information. Software is designed toprocess information—that is its entire function—so function hereis integrally linked with communication, information andknowledge.In Canada, there has been no success in invalidating theCopyright Act against the freedom of communication in Article 2of the Charter of Rights and Freedoms. It has been suggestedthat in Canada the idea /expression dichotomy is not sufficient toensure free speech, especially in the absence of a broad based fairuse doctrine.170 Canada, like Australia and the UK, has a muchmore limited and specific fair dealing doctrine.171 In Australia,there seems little doubt that the implied freedom of politicalcommunication could be invoked in a challenge to the CopyrightAct in an instance of political discourse. Short of that, free speechconsiderations will need to be advocated through theinterpretation of copyright law. In Grain Pool,172 Judge Kirbyexplained:The protection of intellectual property rights must beafforded in a constitutional setting which upholds othervalues of public good in a representative democracy. Inthe United States the relevant head of constitutionalpower has been viewed as containing in-built limitationsmany of which are derived from the competingconstitutional objective of public access to information. InAustralia, the constitutional setting is different but the170. See Fewer, supra note 19, at 217-19. There are also issues raised by theCanadian notion of state action, which does not easily extend to common/judge madelaw and arguably non-government actors such as copyright owners.171. Copyright Law Reform Committee, supra note 60, at Part 1, Exceptions to theExclusive Rights of Copyright Owners (1998) (Austl.) (suggesting that Australia adopta U.S. styled fair use provision).172. [2000] H.C.A. 14 (Austl.).

existence of competing constitutional objectives, expressand implied, is undoubted. 173The message from this passage is that intellectual propertyrights are subject to constitutional limits; yet, can it be suggestedthat fundamental democratic principles such as public access toinformation might act to influence the ultimate shape of copyrightor information law?There is a deeper question concerning the social, as opposed toeconomic, aspect of innovation in this entire debate. It revolvesaround a fundamental (engineering) principle of diversity inknowledge creation or distributed intelligence that is the hallmarkof a democratic society. The notion of “diversity” stems from thecases of Associated Press v. United States174 and TurnerBroadcasting Systems Inc. v. FCC,175 where the United StatesSupreme Court explained: “The basic tenet of nationalcommunication policy is that the widest possible dissemination ofinformation from diverse and antagonistic sources is essential tothe welfare of the public.”176 This ethic also underpins the FirstAmendment.177 A fundamental engineering principle of socialand cultural communication pathways is that of diversity—monopoly of thought will not assist a pluralistic and tolerantsociety fuelled by distributed intelligence. For now the challengeby the First Amendment to the Copyright Act or the Patent Acthas yet to be fully realised. This is not to say the arguments willnever prosper. The assertion of free and open discursiveId.173. Id. at n.218.174. 326 U.S. 1, 20 (1944).175. 512 U.S. 622, 662-64 (1994).176. Id. at 663.177. See Whitney v. California, 274 U.S. 357, 375 (1927) (Brandeis, J., concurring).Those who won our independence believed that the final end of the State wasto make men free to develop their faculties; and that in its government thedeliberative forces should prevail over the arbitrary. They valued liberty bothas an end and as a means. They believed liberty to be the secret of happinessand courage to be the secret of liberty. They believed that freedom to think asyou will and to speak as you think are means indispensable to the discoveryand spread of political truth; that without free speech and assemblydiscussion would be futile; that with them, discussion affords ordinarilyadequate protection against the dissemination of noxious doctrine; that thegreatest menace to freedom is an inert people; that public discussion is apolitical duty; and that this should be a fundamental principle of theAmerican government.

frameworks that will allow diversity of opinion is of paramountimportance to any society wishing to encourage independentthoughts and distributed (as opposed to centrally created)intelligence.And as digital property is much more essential tocommunication than real property has ever been, the culturalaspect of its construction will continue to be highlighted. Atpresent, though, these arguments are not eagerly received. Themore fruitful area for debate is in relation to the general lawdefining and refining digital property.III. THE DEFINING ASPECTS OF GENERAL (AS OPPOSED TOCONSTITUTIONAL) STATUTORY OR COMMON LAWWhile a digital property right might be constitutionalpursuant to the Copyright and Patent Clause or the CommerceClause and survive First Amendment scrutiny in the UnitedStates, Article 2 in Canada, or the implied right to free politicalspeech in Australia, it will still take definition from otherprinciples of ordinary or general (as opposed to constitutional)law. In particular, digital property rights must “fit”178 with otherfundamental principles of general law, some of which exhibit aconstitutional-like presence in terms of tradition and content.Much of the specific definition of intellectual property lawcomes through statutes. My aim here is not to rehash those lawsin detail but to look more at laws that limit those positive grantsof property. There are a number of principles of law that act toencumber digital property rights with an obligation of “diversity.”This notion is the same as explained above in relation to freespeech. Principles of general law, like antitrust/competition,copyright fair use doctrine (e.g., reverse engineering of software)and copyright misuse doctrine act to prevent extrememonopolisation or control of informational products by any oneperson. These principles of general law are dynamic in the sensethat they capture the imagination and confidence of the courtsand have been much more successful in harnessing the untoldpower of digital property rights. We must look more closely at thisarea of law, as much of the construction of digital property asplayed out in the courts is being undertaken in this space.178. See Burnie Port Auth. v. Gen. Jones Pty. Ltd. (1994) 179 C.L.R. 520 (Austl.);Ronald Dworkin, Law’s Empire (Harvard University Press 1986).

These laws also make an important statement aboutprinciples in our legal system and the notion of constitutionalism.Principles of law provide us with guidelines forregulating/mediating an exercise of power. They come in thetraditional constitutional variety (relating to an exercise ofgovernment power: vertical constitutionalism), but also in a muchmore “private sphere” common law or statutory variety: horizontalconstitutionalism. Examples of the latter are common lawprinciples such as reasonable care, unconscionability and unjustenrichment or statutory principles such as antitrust and fairuse/dealing. While this group of principles is not seen asconstitutional law in the strict sense, they represent principles ofconstitutionalism (guidelines for regulating power relations)179and will be increasingly important in this form ofconceptualization to the development of a theory of digitalconstitutionalism (discussed below). In short, non-constitutionalprinciples of ordinary or general law as laid out in statutes orjudicial decisions will bring definition to digital property. Theinterstices of ordinary law and the Dworkinian notion of “fit” bringmuch understanding to the defining process of digital property.180A. Antitrust/Competition LawThere is an inherent tension between IP law that creates aform of monopoly and antitrust or competition law that monitorsthe power of monopolies in the market place.181 It is becomingincreasingly apparent that digital property will take a significantaspect of its definition from antitrust law or similar doctrines likecopyright misuse, which is a defence to an action in copyright.Antitrust or competition law, then, is the first fundamentalprinciple to consider. Section 2 of the Sherman Act182 provides179. Michel Foucault, Power/Knowledge (Colin Gordon ed., Pantheon Books 1980)(1972).180. See generally Maggbury Pty. Ltd. v. Hafele Australia Pty. Ltd. [2001] H.C.A.70 (employing restraint of trade doctrine in relation to a contractual obligationconcerning information).181. Alcatel USA, Inc. v. DGI Technologies, Inc., 166 F.3d 772 (5th Cir. 1999)(discussing the notion of copyright and patent misuse as defences to actions forinfringement rather than causes of action in themselves).182. 15 U.S.C. § 2 (1994).Section 2. Every person who shall monopolize, or attempt to monopolize, orcombine or conspire with any other person or persons, to monopolize any partof the trade or commerce among the several States, or with foreign nations,

that a monopolist should not seek to grow or maintain183 amonopoly through anti-competitive conduct. Section 46 of theAustralian Trade Practices Act says that someone possessingsubstantial market power should not take advantage of thatpower to lessen competition, while in Europe an entity should notabuse a dominant position in the market.184Understanding the scope and function of these laws dependson understanding their purpose. There are at least two reasonsfor adopting competition laws:185to facilitate diversity of economic actors in the marketplace—suggesting a number of small or medium playersrather than monopolies and;to ensure economic efficiency leading to an optimisationof consumer welfare—suggesting a market place thatprovides the best economic outcomes for consumersThe first rationale promotes diversity while the second promoteseconomic efficiency, in terms of economic welfare.Many subscribe to the view that the Sherman Act was enactedto protect the existence of small or medium size businesses in theface of monopolies.186 However, under the influence of theChicago School of Law and Economics, antitrust law is nowadaysmore frequently conceptualized in terms of the economic efficiencyof market practices with consumer welfare as an economic issueshall be deemed guilty of a felony, and, on conviction thereof, shall bepunished by fine not exceeding $ 10,000,000 if a corporation, or, if any otherperson, $ 350,000, or by imprisonment not exceeding three years, or by bothsaid punishments, in the discretion of the court.Id.183. In Aspen Skiing Co. v. Aspen Highlands Skiing Corp., the Supreme Courtexplained that “[t]he offense of monopoly under §2 of the Sherman Act has twoelements: (1) the possession of monopoly power in the relevant market and (2) thewillful acquisition or maintenance of that power as distinguished from growth ordevelopment as a consequence of a superior product, business acumen, or historicaccident.” 472 U.S. 585, 596 (1985) (quoting United States v. Grinnell Corp., 384 U.S.563, 570-71 (1966)).184. See Trade Practices Act, 1974, § 46 (Austl.).185. See David McGowan, Innovation, Uncertainty and Stability in Antitrust Law,16 Berkeley Tech. L.J. 729, 741-64 (2001) (discussing “some of antitrust’s purposes”).186. Herbert Hovenkamp, Federal Antitrust Policy 47-76 (2d ed. 1999) (arguingthis view is supported by more recent amendments to the Sherman Act and publicchoice theory of the role of interest groups in the legislative process); George J. Stigler,The Origin of the Sherman Act, 14 J. of Legal Stud. 1, 1-8 (1985).

eing the ultimate criterion.187 The economic efficiency approachat points seems to ignore political and social aspects of themarket.188 In Australia, section 2 of the Trade Practices Actprovides: “The object of this Act is to enhance the welfare ofAustralians through the promotion of competition and fair tradingand provision for consumer protection.”189David McGowan has recently written190 about the competingrationales of U.S. antitrust law in light of the Microsoft191decision. He ponders whether the decision is really aboutguaranteeing the existence of smaller firms and competition in themarket place or whether it is about total economic efficiencyanalysed through producer and consumer equations and evidencedin optimal costs and pricing.192 He argues that Judge Jackson inthe Microsoft case seemed to want to invoke both notions.193 Hequestions whether “innovation,” as invoked in antitrust reasoning,is a pursuit of many small players in the market or the pursuit ofbest possible outcomes in terms of economic welfare.194So what is it that we really want from antitrust law—diversity or optimal consumer welfare? The two are notnecessarily congruent. If we choose diversity as the touchstone,the exploitation of intellectual property rights will be defined inthis light and we may see courts more willing to facilitate abroader range of access to informational products than we wouldunder an economic efficiency approach. And if we marry a culturaltheory of intellectual property law with a cultural theory ofantitrust law, we are almost certain to see differently constructeddigital property rights than if economic theory were the187. See Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis ofAntitrust Principles and Their Application, 8-11, 51-57, 94-137 (2d ed. 2000) (calling fora multi-dimensional approach, yet arguing for the primary importance of consumerwelfare over rivalry of small business—although economic theory is said to be limited);Robert H. Bork, The Antitrust Paradox 15-134 (1978).188. See United States v. Aluminum Co. of Am., 148 F. 2d 416, 428-29 (2d Cir.1945); United States v. Associated Press, 52 F. Supp. 362, 370-72 (S.D.N.Y. 1943); seealso Areeda & Hovenkamp, supra note 187, at 97-115; Bork, supra note 187, at 50-56;Louis B. Schwartz, “Justice” and Other Non Economic Goals of Antitrust, 127 U. Pa. L.Rev. 1076 (1979).189. Trade Practices Act, 1974, § 46.190. See McGowan, supra note 185.191. United States v. Microsoft Corp., 87 F. Supp. 2d 30 (D.D.C. 2000).192. See McGowan, supra note 185, at 766-76.193. Id.194. Id.

motivating force.While antitrust or competition law may be seen by many tonow rest in the lap of economic analysis, there can be little doubtthat growing proprietary rights in digital architecture, such assoftware, which forms the basis of modern communication (formand content), are demanding that a diversity of communicationpathways be opened.195 The role of antitrust law in ensuringdiversity of information and knowledge (in the mould of the FirstAmendment) is clearly articulated in the following passage fromthe Associated Press case:The First Amendment, far from providing an argumentagainst application of the Sherman Act, here providespowerful reasons to the contrary. That Amendment restson the assumption that the widest possible disseminationof information from diverse and antagonistic sources isessential to the welfare of the public, that a free press is acondition of a free society. Surely a command that thegovernment itself shall not impede the free flow of ideasdoes not afford non-governmental combinations a refugeif they impose restraints upon that constitutionallyguaranteed freedom. Freedom to publish means freedomfor all and not for some. Freedom to publish isguaranteed by the Constitution, but freedom to combineto keep others from publishing is not. Freedom of thepress from governmental interference under the FirstAmendment does not sanction repression of that freedomby private interests.196This passage, which some might suggest is outdated and othersmight characterise as timeless, highlights that welfare of thepublic is a social and not just an economic issue and that private195. See, e.g., Microsoft, 87 F. Supp. 2d 30; Melway Publishing Pty. Ltd. v. RobertHicks Pty. Ltd. [2001] H.C.A. 13 (Austl.).196. Associated Press v. United States, 326 U.S. 1, 20 (1945); see also UnitedStates v. Associated Press, 52 F. Supp. 362, 372 (S.D.N.Y. 1943). There, the districtcourt stated:Id.That interest [of the news industry to present a diversity of opinion] is closelyakin to, if indeed it is not the same as, the interest protected by the FirstAmendment; it presupposes that right conclusions are more likely to begathered out of a multitude of tongues, than through any kind ofauthoritative selection. To many this is, and always will be, folly: but wehave staked upon it our all.

corporations in their exercise of power are subject to constitutionallike principles (another aspect of the notion of digitalconstitutionalism) of diversity rooted in antitrust law. Here liesan extremely important point. Access and user rights to—consumption of—information and the ensuing construction ofknowledge should never be the domain of one entity, public orprivate. Competition law has the potential to implement such aprinciple and, if it is invoked in this manner, it will be central tothe definition of digital property.1. “Refusal to Deal”One of the crucial issues that does much to define the scope ofdigital property is the extent to which someone can refuse tosupply their property to a particular customer. This issue isusually much more delicate where the supplier has an intellectualproperty right (copyright or patent) to enforce the refusal tosupply. In an environment where private corporations are makingthe highways and languages of the digital world, it becomesfrightening to think they have the power to refuse entry to digitaldiscourse.“Refusal to deal”197 has been considered in a number ofcases.198 In essence, it means that the supplier wishes todiscriminate against a particular purchaser, normally for strategicbusiness reasons, but also potentially for discrimination on thebasis of ideology. Generally, as a matter of contract law, we canrefuse to sell tangible property to someone so long as we do notinfringe discrimination laws concerning things such as race, sexand religion. That kind of strategic business discrimination isgenerally seen as part of the ethical cycle of business, and isconfirmed in Melway Publishing Pty. Ltd. v. Robert Hicks Pty.Ltd.199 However, there is a fine line as to when this becomes197. That is, the right to refuse to sell a product to someone.198. Compare Otter Tail Power Co. v. United States, 410 U.S. 366 (1973) (statingthat in terms of non-intellectual property protected products in a continuingrelationship, refusal to sell may be an antitrust violation) and Aspen Skiing Co. v.Aspen Highlands Skiing Corp., 472 U.S. 585 (1985) with Intergraph Corp. v. IntelCorp., 195 F.3d 1346, 1362 (Fed. Cir. 1999) (stating that a refusal to license a patent orsell a patented item without more could not be an antitrust violation) and In re Indep.Serv. Orgs. Antitrust Litig., 203 F.3d 1322, 1329 (Fed. Cir. 2000); cf., Image TechnicalServ., Inc. v. Eastman Kodak Co., 125 F.3d 1195 (9th Cir. 1997) (holding that such arefusal may be an antitrust violation).199. [2001] H.C.A. 13(Austl.).

unjustifiable anti-competitive conduct. In the United States, incases of a continuous relationship, this will be more likely toinfringe.200The recent Independent Service Organizations AntitrustLitigation201 decision highlights some of these issues. In thatcase, Xerox made high volume copiers and had adopted a policy ofrefusing to sell parts that were unique to its “Series 10” copiers toindependent service organizations (“ISOs”), including the companyCSU, unless they were also end users of such copiers.202 Thepolicy was expanded in 1987 to include all new products and theolder “Series 9” copiers.203 In 1989, the enforcement of the policywas tightened, which resulted in Xerox cutting off CSU’s ability todirectly purchase restricted parts.204 Xerox “also implemented an‘on-site, end user verification’ procedure to confirm that the partsordered by certain ISOs or their customers were actually for theirend user use.”205 That policy was applied to CSU, one of the moresuccessful ISOs.206In order to keep servicing Xerox equipment, CSU: used partsfrom used Xerox equipment; used parts obtained from other ISOsand used parts purchased from customers.207 In addition, CSUalso obtained parts from Rank Xerox, a European Xerox affiliate,until Xerox stopped that practice.208 Xerox settled an antitrustlawsuit with some ISOs and agreed to suspend its restrictive partspolicy and to license its diagnostic software for agreed uponperiods.209 CSU opted out of that settlement and filed suit,alleging that Xerox violated the Sherman Act by setting the priceson its patented parts much higher for ISOs than for end users inan attempt to force ISOs to raise prices.210 By that, the alleged200. See Otter Tail Power, 410 U.S. 366; Aspen Skiing, 472 U.S. 585.201. 203 F.3d 1322 (Fed. Cir. 2000).202. Id. at 1324203. Id.204. Id.205. Id.206. Id. Compare this case where Xerox said we will not sell you our parts unlessyou own the machine to the scenario in Kodak where Kodak said you cannot buy themachine unless you buy the service, which resulted in a finding of tying under section 1of the Sherman Act (no contract should unreasonably restrain trade). See EastmanKodak Co. v. Image Technical Servs., Inc., 504 U.S. 451 (1992).207. In re Indep. Serv. Orgs. Antitrust Litig., 203 F.3d at 1324.208. Id.209. Id.210. Id.

violation would then eliminate ISOs, including CSU, ascompetitors in the market of high-speed copiers and printers.Xerox responded with patent and copyright infringementcounterclaims and also contested CSU’s antitrust claims becausethey relied on injuries that they alleged were solely caused by therefusal to sell and license patented parts and copyrightedsoftware, which Xerox considered lawful.211In dismissing the antitrust claims against Xerox, the courtexplained:Intellectual property rights do not confer a privilege toviolate the antitrust laws. “But it is also correct that theantitrust laws do not negate the patentee’s right toexclude others from patent property . . . .”A patent alone does not demonstrate market power . . . .The patentee’s right to exclude, however, is not withoutlimit. As we recently observed in Glass EquipmentDevelopment Inc. v. Besten, Inc., a patent owner whobrings suit to enforce the statutory right to exclude othersfrom making, using, or selling the claimed invention isexempt from the antitrust laws, even though such a suitmay have an anticompetitive effect, unless theinfringement defendant proves one of two conditions.First, he may prove that the asserted patent was obtainedthrough knowing and willful fraud within the meaning ofWalker Process Equipment, Inc. v. Food Machinery &Chemical Corp. Or he may demonstrate that theinfringement suit was a mere sham to cover what isactually no more than an attempt to interfere directlywith the business relationships of a competitor. Here,CSU makes no claim that Xerox obtained its patentsthrough fraud in the Patent and Trademark Office; theWalker Process analysis is not implicated.. . . .. . . We therefore will not inquire into his subjectivemotivation for exerting his statutory rights, even though211. Id.

his refusal to sell or license his patented invention mayhave an anticompetitive effect, so long as thatanticompetitive effect is not illegally extended beyond thestatutory patent grant. It is the infringement defendantand not the patentee that bears the burden to show thatone of these exceptional situations exists and, in theabsence of such proof, we will not inquire into thepatentee’s motivations for asserting his statutory right toexclude. Even in cases where the infringement defendanthas met this burden, which CSU has not, he must thenalso prove the elements of the Sherman Act violation.212The court then turned to the issue of copyright, stating:The Copyright Act expressly grants a copyright owner theexclusive right to distribute the protected work by“transfer of ownership, or by rental, lease, or lending.”“[T]he owner of the copyright, if [it] pleases, may refrainfrom vending or licensing and content [itself] with simplyexercising the right to exclude others from using [its]property.”The Supreme Court has made clear that the propertyright granted by copyright law cannot be used withimpunity to extend power in the marketplace beyondwhat Congress intended. The Court has not, however,directly addressed the antitrust implications of aunilateral refusal to sell or license copyrightedexpression.The Tenth Circuit has not addressed in any publishedopinion the extent to which the unilateral refusal to sellor license copyrighted expression can form the basis of aviolation of the Sherman Act. We are therefore left todetermine how that circuit would likely resolve the issue;the precedent of other circuits is instructive in thatconsideration. The Fourth Circuit has rejected a claim ofillegal tying, supported only by evidence of a unilateraldecision to license copyrighted diagnostic software tosome but not to others. In reaching this conclusion, the212. In re Indep. Serv. Org. Antitrust Litig., 203 F.3d 1322, 1325-28 (Fed. Cir.2000) (citations omitted).

court recognized the copyright owner’s exclusive right to“sell, rent, lease, lend, or otherwise distribute copies of acopyrighted work,” and concluded that “Section 1 of theSherman Act does not entitle ‘a purchaser . . . to buy aproduct that the seller does not wish to offer for sale.’”Perhaps the most extensive analysis of the effect of aunilateral refusal to license copyrighted expression wasconducted by the First Circuit in Data General Corp. v.Gruman Systems Support Corp. There, the court notedthat the limited copyright monopoly is based on Congress’empirical assumption that the right to “exclude othersfrom using their works creates a system of incentives thatpromotes consumer welfare in the long term byencouraging investment in the creation of desirableartistic and functional works of expression . . . . Wecannot require antitrust defendants to prove and reprovethe merits of this legislative assumption in every casewhere a refusal to license a copyrighted work comesunder attack.” The court went on to establish as a legalstandard that “while exclusionary conduct can include amonopolist’s unilateral refusal to license a copyright, anauthor’s desire to exclude others from use of itscopyrighted work is a presumptively valid businessjustification for any immediate harm to consumers.” Theburden to overcome this presumption was firmly placedon the antitrust plaintiff. The court gave no weight toevidence showing knowledge that developing aproprietary position would help to maintain a monopolyin the service market in the face of contrary evidence ofthe defendant’s desire to develop state-of-the-artdiagnostic software to enhance its service and consumerbenefit.As discussed above, the Ninth Circuit adopted a modifiedversion of this Data General standard. Both courtsagreed that the presumption could be rebutted byevidence that “the monopolist acquired the protection ofthe intellectual property laws in an unlawful manner.”The Ninth Circuit, however, extended the possible meansof rebutting the presumption to include evidence that the

defense and exploitation of the copyright grant wasmerely a pretextual business justification to maskanticompetitive conduct . . . . This is in reality asignificant departure from the First Circuit’s centralpremise that rebutting the presumption would be anuphill battle and would only be appropriate in those rarecases in which imposing antitrust liability is unlikely tofrustrate the objectives of the Copyright Act.We believe the First Circuit’s approach is moreconsistent with both the antitrust and the copyright lawsand is the standard that would most likely be followed bythe Tenth Circuit in considering the effect of Xerox’sunilateral right to refuse to license or sell copyrightedmanuals and diagnostic software on liability under theantitrust laws. We therefore reject CSU’s invitation toexamine Xerox’s subjective motivation in asserting itsright to exclude under the copyright laws for pretext, inthe absence of any evidence that the copyrights wereobtained by unlawful means or were used to gainmonopoly power beyond the statutory copyright grantedby Congress. In the absence of such definitive rebuttalevidence, Xerox’s refusal to sell or license its copyrightedworks was squarely within the rights granted byCongress to the copyright holder and did not constitute aviolation of the antitrust laws.213Both of these passages highlight the delicate interplaybetween intellectual property law, antitrust law and theconstitutionally enumerated grant of power. The ambit of thecopyright or patent is defined by the meaning of antitrust law, butantitrust law is also confined by the ambit of copyright andpatent. One may refuse access to or use of an informationalproduct or digital property, even where someone is willing to paymarket price, so long as one does not engage in anti-competitiveconduct. The questions then return to what anti-competitiveconduct means and what is the desired or legislated role ofantitrust law.Intergraph Corp. v. Intel Corp.214 is another case addressing213. Id. at 1328-29 (citations omitted).214. 195 F.3d 1346 (Fed. Cir. 1999).

efusal to deal. In that case, Intel, a manufacturer of computermicroprocessors, supplied Intergraph, an original equipmentmanufacturer (OEM), with microprocessors for their computerworkstations.215 Intel saw Intergraph as a special customer andprovided Intergraph with various special benefits, includingproprietary information and products, under non-disclosureagreements.216 Through a series of events in November 1997,Intergraph sued Intel for infringement of the certain patents thatit held.217 In this litigation, Intergraph also sought an injunctionto stop Intel from cutting off the supply of benefits toIntergraph.218 Intel opposed this motion and Intergraph amendedtheir claim to say that Intel was in violation of antitrust laws,basically on the basis that it refused to deal its information toIntergraph, even though Intergraph was suing it on other groundsfor patent infringement.219 The question was, in essence, is itlegitimate to refuse to supply someone with your informationalvalue in the face of suit by them on other grounds?The district court enjoined Intel from refusing to supplyinformation and Intel appealed, arguing that no law required it togive such special benefits, including its trade secrets, proprietaryinformation, intellectual property, pre-release products, allocationof new products and other preferences, to an entity that was suingit on charges of multiple wrongdoings.220The court noted that Intergraph and Intel operated indifferent markets and were not competitors.221 While Intel mighthold a large share of the market for microprocessors, it could notbe held liable for growing its monopoly through anti-competitiveconduct by a non-competitor or consumer.222 Intergraph alsoclaimed that Intel’s information was an essential facility and thatit should be allowed access to it.223 The essential facility doctrineprovides an obligation to deal/supply but usually relates to coretangible infrastructure, like railway lines or electricity wires. In215. Id. at 1349.216. Id. at 1350.217. Id.218. Id.219. Id.220. Intergraph Corp., 195 F.3d at 1350.221. Id. at 1352-56.222. Id. at 1353.223. Id. at 1356.

the information age there have been a number of claims that coreinfrastructure of the information society is an essential facilityand cannot be closed off from use.224The court explained the notion of an essential facility asfollows:The “essential facility” theory of Sherman Act violationstems from United States v. Terminal RR Ass’n, whereina group of railroads formed an association that controlledthe railroad terminals, bridges, and switching yardsserving the City of St. Louis. The Court held that thisassociation was formed for an anticompetitive purpose,that the railroad terminals, bridges, and yards werefacilities essential to competing railroads, and thatsection 1 of the Sherman Act was violated.The district court found that “the Advance ChipsSamples and advance design and technical informationare essential products and information necessary forIntergraph to compete in its markets.” Reasoning that“[t]he antitrust laws impose on firms controlling anessential facility the obligation to make the facilityavailable on non-discriminatory terms,” the court heldthat Intel’s action in withdrawing these benefits violatedthe Sherman Act. As authority the district court citedOtter Tail Power Co. v. United States and MCICommunications Corp. v. American Telephone &Telegraph Co.. . . .In MCI Communications, the court enumerated theelements of liability under the “essential facilities” theoryas “(1) control of the essential facility by a monopolist; (2)a competitor’s inability practically or reasonably toduplicate the essential facility; (3) the denial of the use ofthe facility to a competitor; and (4) the feasibility ofproviding the facility.” The courts have well understoodthat the essential facility theory is not an invitation to224. See eBay, Inc. v. Bidder’s Edge, Inc., 100 F. Supp. 2d 1058 (N.D. Cal. 2000);Cyber Promotions, Inc. v. Am. Online, Inc., 948 F. Supp. 436 (E.D. Pa. 1996);Compuserve, Inc. v. Cyber Promotions, Inc., 962 F. Supp. 1015 (S.D. Ohio 1997).

demand access to the property or privileges of another, onpain of antitrust penalties and compulsion; thus thecourts have required anticompetitive action by amonopolist that is intended to “eliminate competition inthe downstream market.”225Ultimately, the court concluded this was not an essential facilityscenario as Intel and Intergraph were not competitors and thenmoved on to reject the notion that Intel had done something wrongin refusing to supply its informational products to Intergraph:Other than as a remedy for illegal acts, the antitrust lawsdo not compel a company to do business with anyone—customer, supplier, or competitor.. . . [I]t is well established that “in the absence of anypurpose to create or maintain a monopoly, the [Sherman]act does not restrict the long recognized right of a traderor manufacturer engaged in an entirely private business,freely to exercise his own independent discretion as toparties with whom he will deal.”. . . .A “refusal to deal” may raise antitrust concerns when therefusal is directed against competition and the purpose isto create, maintain, or . . . enlarge a monopoly. Forexample, in Lorain Journal Co. v. United States, the onlynewspaper in town refused to sell newspaper advertisingto persons who also advertised on a competing radiostation; this was held to be an attempt to . . . monopolizethe mass dissemination of all news and advertising, andto violate the Sherman Act.226Claims of leveraging (using monopoly power to gain a competitiveadvantage in another market through illegitimate conduct),coercive reciprocity and tying (based on the allegation that Intelwould supply the information if Intergraph settled the law suit)were also rejected by the court.227Finally, the court more closely considered the issue ofintellectual property rights. The court explained:225. Intergraph Corp., 195 F.3d at 1356-57 (citations omitted).226. Id. at 1357-58 (citations omitted).227. Id. at 1359-60.

In response to Intel’s argument that its proprietaryinformation and pre-release products are subject tocopyright and patents, the district court observed thatIntel’s intellectual property “does not confer upon it aprivilege or immunity to violate the antitrust laws.” Thatis of course correct. But it is also correct that theantitrust laws do not negate the patentee’s right toexclude others from patent property. The patent andantitrust laws are complementary, the patent systemserving to encourage invention and the bringing of newproducts to market by adjusting investment-based risk,and the antitrust laws serving to foster industrialcompetition. The patent and antitrust laws serve thepublic in different ways, both of importance to the nation.. . . .In Image Technical Services the Ninth Circuit reportedthat it had found “no reported case in which a court hadimposed antitrust liability for a unilateral refusal to sellor license a patent or copyright.” Nor have we.. . . However, as we have stated, the owner of proprietaryinformation has no obligation to provide it, whether to acompetitor, customer, or supplier. Precedent makes clearthat a customer who is dependent on a manufacturer’ssupply of a component can not on that ground force theproducer to provide it; there must also be ananticompetitive aspect invoking the Sherman Act.228The message from this case is that refusal to deal or supplyinformation/intellectual property to a competitor or consumer isnot of itself an antitrust violation. Consumers will generally havea difficult time raising successful antitrust violations, as anticompetitiveconduct in competition is an important requirement.Where the dispute involves competitors, the refusal to supply, if itamounts to anti-competitive conduct, will raise a more seriousissue. This case also confirms that the essential facility doctrinewill be of limited application and will not readily provide a right toaccess informational architecture.229228. Id. at 1662-63 (citations omitted).229. See, e.g., Trade Practices Act, 1974, Part IIIA (Austl.); Hamersley Iron Pty.Ltd. v. NCC [1999] F.C.A. 1370; Eastman Kodak Co., 504 U.S. 451; Aspen Skiing, 472

ISO and Intergraph can be contrasted with Radio TelefisEireann v. European Commission (Magill), a decision of theEuropean Court of Justice (ECJ).230 In Magill, due to the specificcopyright laws in place, television stations were able to claimcopyright in the listing of weekly programs in Ireland. Thetelevision stations published their own station specific as opposedto comprehensive weekly television guides and released daily andweekend details of programs to daily newspapers.231 Magillattempted to publish a comprehensive weekly television guide butwas enjoined from doing so on the motion of the televisionstations.232 Magill sought a declaration that in refusing to licencethe copyrighted material the stations had engaged in antitrustviolations or in European terms abused their dominant position inthe market place.233 The ECJ agreed, opining:In the present case, the conduct objected to is theappellants’ reliance on copyright conferred by nationallegislation so as to prevent Magill—or any otherundertaking having the same intention—from publishingon a weekly basis information (channel, day, time andtitle of programmes) together with commentaries andpictures obtained independently of the appellants.Among the circumstances taken into account by theU.S. 585; In re Indep. Serv. Orgs. Antitrust Litig., 203 F.3d 1322; Intergraph Corp. v.Intel Corp., 3 F. Supp. 2d 1255 (N.D. Ala. 1998), vacated by 195 F.3d 1346 (Fed. Cir.1999); Eastman Kodak Co. v. Image Technical Servs., Inc., 125 F.3d 1195 (9th Cir.1997); Data Gen. Corp. v. Grumman Sys. Support Corp., 36 F.3d 1147 (1st Cir. 1994);MCI Communications v. AT&T, 708 F.2d 1081 (7th Cir. 1983), cert. denied, 464 U.S.891 (1983); Byars v. Bluff City News Co., 609 F.2d 843 (6th Cir. 1980); Hecht v. Pro-Football Inc., 570 F.2d 982 (D.C. Cir. 1977), cert. denied, 436 U.S. 956 (1978); see alsoPhillip Areeda, Essential Facilities: An Epithet in Need of Limiting Principles, 58 ABAAntitrust L.J. 841 (1990); Allen Kezsbom & Alan Goldman, No Shortcut to AntitrustAnalysis: The Twisted Journey of the “Essential Facilities” Doctrine, 1 Colum. Bus. L.Rev. 1 (1996); David McGowan, Regulating Competition in the Information Age:Computer Software as an Essential Facility Under the Sherman Act, 18 HastingsComm. & Ent. L.J. 771 (1996); W. Greg Paciak, Essential Facilities Doctrine:Intergraph Corp. v. Intel Corp., 14 Berkeley Tech. L.J. 323 (1999).230. See Radio Telefis Eireann v. European Comm’n, 1995 E.C.R. I-743; 4 C.M.L.R.718 (1995); see also TRIPS, supra note 5, at art. 40; Inge Govaere, The Use and Abuseof Intell. Prop. Rts., in E.C. Law 3-12 (1996); Irini A. Stamatoudi, The Hidden Agendain Magill and Its Impact on New Technologies, 1 J. World Intell. Prop. 153 (1998).231. Radio Telefis, 4 C.M.L.R. at para. 3.232. Id. at para. 7.233. Id. at para. 8.

Court of First Instance in concluding that such conductwas abusive was, first, the fact that there was, accordingto the findings of the Court of First Instance, no actual orpotential substitute for a weekly television guide offeringinformation on the programmes for the week ahead. Onthis point, the Court of First Instance confirmed theCommission’s finding that the complete lists ofprogrammes for a 24-hour period—and for a 48-hourperiod at weekends and before public holidays—publishedin certain daily and Sunday newspapers, and thetelevision sections of certain magazines covering, inaddition, “highlights” of the week’s programmes, wereonly to a limited extent substitutable for advanceinformation to viewers on all the week’s programmes.Only weekly television guides containing comprehensivelistings for the week ahead would enable users to decidein advance which programmes they wished to follow andarrange their leisure activities for the week accordingly.The Court of First Instance also established that therewas a specific, constant and regular potential demand onthe part of consumers.Thus the appellants—who were, by force ofcircumstance, the only sources of the basic information onprogramme scheduling which is the indispensable rawmaterial for compiling a weekly television guide—gaveviewers wishing to obtain information on the choice ofprogrammes for the week ahead no choice but to buy theweekly guides for each station and draw from each ofthem the information they needed to make comparisons.The appellants’ refusal to provide basic information byrelying on national copyright provisions thus preventedthe appearance of a new product, a comprehensive weeklyguide to television programmes, which the appellants didnot offer and for which there was a potential consumerdemand. Such refusal constitutes an abuse under heading(b) of the second paragraph of Article 86 of the Treaty.Second, there was no justification for such refusaleither in the activity of television broadcasting or in thatof publishing television magazines.

Third, and finally, as the Court of First Instance alsoheld, the appellants, by their conduct, reserved tothemselves the secondary market of weekly televisionguides by excluding all competition on that market sincethey denied access to the basic information which is theraw material indispensable for the compilation of such aguide.234The case is significant because it held that the copyright ownerwas not allowed to refuse to supply information, thereby makingthe TV listings tantamount to an essential facility.235 Thisdecision of the ECJ, then, provides a more robust view in thecontext of information as to when it is legitimate to refuse to deal.Melway236 provides yet another perspective on this notion ofrefusal to deal. Melway’s, who produced a street directory forMelbourne, Australian, refused to supply the respondent, awholesaler of motor vehicle parts and accessories, and a formerauthorised distributor, supplies of directories.237 The respondentalleged a breach of section 46 of the Trade Practices Act, whichprohibits “taking advantage” of substantial market power tolessen competition.238 The court confirmed the relevantfunctional market was the wholesale and retail market for streetdirectories in Melbourne and that Melway’s held in 80–90% of theretail market share for street directories.239 Melway’s, which hadbeen marketing the directories since 1966, had developed astrategic method for distributing and selling the directories.240The respondent had formerly been involved as a distributor buthad that right terminated pursuant to the distributorshipagreement.241The majority of the Australian High Court held that Melway’shad the right to determine how to distribute their informationalproduct.242 In the absence of any evidence that they had intended234. Id. at paras. 51-56 (citations and footnotes omitted).235. Id. at paras. 76-88.236. [2001] H.C.A. 13 (Austl.).237. Id.238. Id. at para. 1.239. Id. at para. 80.240. Id. The way they distribute could also be seen as an architectural constrainton the way they do business—like code—by slowing down copyists or, in other words,implementing their copyright.241. Id. at para. 87.242. [2001] H.C.A. 13 at para. 80.

to oust competitors, there was nothing wrong with what they haddone—in fact, it was simply a normal business decision.243 Theeconomic effect of Melway’s action and ultimate consumer welfare(from an economic sense) seemed to be in harmony. Melway’s wassuccessful in having their action vindicated.Judge Kirby’s dissent was geared toward guaranteeing andensuring diversity of suppliers and access to the informationalproduct. He explained:The respondent’s primary submission was that,properly analysed, Queensland Wire stands for only onerelevant legal proposition applicable to this case. Thiswas that “take advantage of” means no more than “use.”In determining whether a corporation has “take[n]advantage of “its market power for a proscribed purpose,the respondent argued that all that was necessary wasproof that, as a matter of fact, the corporation, havingsuch power, had refused supply for a proscribed purpose.Upon this view, it was unnecessary to pose hypotheticalquestions (sometimes difficult to resolve) as to whethersuch corporation could or would, acting rationally, haveengaged in the forbidden conduct if it were subject toeffective competition. My own opinion is that this iscorrect analysis of s 46(1) of the Act. I also consider thatit is what the decision in Queensland Wire stands for as abinding principle of law. There is nothing in thelanguage of the section itself that obliges theascertainment of an answer to a hypothetical question.If, as was held in Queensland Wire, “take advantage of”means no more than “use”, that presents a purely factualquestion to be answered. In short, if the supplier enjoys asubstantial degree of market power, the grant or refusalof supply is necessarily, as a matter of fact, takingadvantage of (sc “using”) such market power. It is doingso because the power to grant, or refuse, supply is part of243. Id. at para 67. The majority explained:But it does not follow that because a firm in fact enjoys freedom from competitiveconstraint, and in fact refuses to supply a particular person, there is a relevantconnection between the freedom and the refusal. Presence of competitive constraintmight be compatible with a similar refusal, especially if it is done to secure businessadvantages which would exist in a competitive environment.Id.

the power substantially to control the market.. . . .However, if I am wrong in the foregoing, it isappropriate to consider the alternative approach on thefooting that the hypothetical question forms part of therule for which Queensland Wire stands. If this is done, itwas certainly open to the majority judges in the FederalCourt to conclude that the impugned conduct on the partof the appellant involved its taking advantage of itsmarket power, in the sense that, acting rationally, itwould and could not (but for that power) have acted as itdid. Specifically, it would and could not have refused therespondent’s offer to purchase between 30,000 and 50,000copies of the Melbourne directory each year.244Judge Kirby’s judgement is to the effect that anyone who holdsmarket power and refuses to supply is taking advantage of thatmarket power. This is very much a diversity (or nondiscrimination)argument that requires the information to bedealt. It is tantamount to a right of access to information. JudgeKirby was not just looking at economic transaction costs but alsothe social innovation cost, remembering innovation has a social aswell as an economic aspect. The delivery of this information needsa diversity of pathways. The monopolisation of the delivery ofinformation to people is not good for them—socially. To thisextent, Judge Kirby perhaps showed a social approach tocompetition law and the notion of welfare.Another interesting issue arises from this case. Melway’sowns an informational product in which they claimed copyright.This case, in essence, defines the right of the copyright owner—the right to distribute or sell the copyright product to whomeverthey wish. Yet in this instance, competition law puts a gloss onthis right. In this case, we are really seeing the definitions ofcompetition and copyright law being conflated—because therefusal to deal is intimately connected with intellectual propertyrights.In summary, the power to make someone distribute aninformation product is contingent upon the power that they hold244. Id. at paras. 109, 112.

as a digital property owner, which will, to some extent, be definedby competition law. Refusal to deal, by itself, is not generallyunlawful, but other factors (consider Magill) will raise the level ofscrutiny; this is a difficult area in which to be confident that youwill not be liable. Judge Kirby’s approach of guaranteeing a rightto deal in core informational products, a little like essentialfacilities doctrine, may suggest that the time is upon us to moreseriously consider access to core informational products as part ofantitrust or competition law.2. The Microsoft CaseThe Microsoft case also highlights how intellectual propertyrights in software may be questioned in light of antitrust law.This case concerned the dominance of the Microsoft operatingsystem, Windows.245 An operating system is software thatallocates and manages the hardware resources of a computer. Theoperating system combined with the hardware is often called theoperating platform upon which many forms of applications(specific kinds of) software are built. To write applications thatwill run on a particular platform or operating system one needs toensure the application can interact with the operating system. Anoperating system will be programmed to allow such interactionthrough what is known as application programming interfaces(APIs). Therefore, to write code or software that will run on anoperating system you will need to know and obey APIs. This needto write applications for specific operating systems means that itis more difficult for a generic application to be written that can berun on all platforms. As Judge Jackson found, Windows was byfar the most dominant operating system in the market, with amarket share in excess of eighty percent depending on how youcalculated the figures.246The emergence of the Internet browser and the developmentof the Java programming language by Sun Microsystems openedthe way for diluting the dominance of the Windows operatingsystem. The browser developed with such speed and popularitythat the potential opened for people to write applications to run onthe browser platform, thereby reducing the control Microsoft had245. When joined with the Intel microprocessor or chip this is commonly referred toas a WinTel platform.246. Microsoft, 87 F. Supp. 2d at 36.

over applications development and also threatening the continuedprimacy of its Windows operating system. Windows would becomemerely a part of the substratum that could be easily substituted ifapplications software could be written to the browser platform.Sun’s programming language, Java, also allowed applications tobe written in such a way that they could be ported from oneplatform to another. Both of those developments threatenedMicrosoft and made it more likely for a competitor such as theopen source operating system, Linux, to be more successful.One way Microsoft took on the challenge was to build up itsmarket share in the internet browser area through its product,Internet Explorer (IE). Its main rival in this area was theNetscape internet browser, which showed the potential to becomea dominant platform upon which to write applications software.Microsoft was also alleged to have tried to make it difficult forJava based programs to run on Windows without modification.247Judge Jackson found that Microsoft breached sections 1 and 2of the Sherman Act.248 They were found to have violated section1—unreasonable contracts in restraint of trade—by tying theirbrowser to the operating system and ensuring through agreementwith OEMs that the two were sold as a unit.249 Microsoft arguedthat Windows and IE were merely one product and there was nounlawful tying. The judge called this a form of technology tyingwhere software code tied the software products together.250 Heheld that there was a clear market for two separate products and,as such, this was a case of tying.251 As far as section 2 wasconcerned, Judge Jackson found that Microsoft’s actions towardsNetscape and Sun were predacious and in breach of the ShermanAct.252 As a remedy, the judge ordered that Microsoft Corporationbe broken up into an applications company and an operatingsystems company.253On appeal, the judge’s findings relating to tying and remedywere questioned.254 However, Judge Jackson’s findings regarding247. Id. at 43.248. Id. at 44-45.249. Id. at 47.250. See id. at 50-51.251. Id.252. Microsoft, 87 F. Supp. 2d at 44.253. United States v. Microsoft Corp., 97 F. Supp. 2d 59, 64-65 (D.D.C. 2000).254. See United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001).

anti-competitive use of monopoly power were not overturned. Thematter has now been sent back to the district court for furtherconsideration, yet how the situation will be remedied remains tobe seen.In summary, this case shows the limitations that can beimposed upon the owner of digital property (copyright in software)in the face of antitrust law.B. Fair UseThe notion of fair use in United States copyright law255 andthe narrower doctrine of fair dealing in UK, Canadian andAustralian copyright law, defines to what extent the digitalproperty owner can exclude others from use of the informationalproduct.256 Fair use and fair dealing allow use without the(personal) permission of the copyright owner; for this reason, somecall it a privilege. In A & M Records, Inc. v. Napster, Inc.257 andReimerdes, fair use arguments were unsuccessfully raised.However, there has been more success in the area of reverseengineering of software.1. Reverse Engineering for InteroperabilityTo ensure the diversity of digital identity we must argue fordiversity in software products.258 Fundamental to softwarediversity is the notion of interoperability, defined in the DMCA as:“the ability of computer programs to exchange information, and ofsuch programs mutually to use the information, which has beenexchanged.”259If a software engineer or developer constructs software thatbecomes the industry standard, that software acts as architecturefor communication and it becomes, in essence, a platform fordiscourse. In order to develop complementary and improvedsoftware products, in most cases, software developers will reverse255. 17 U.S.C. § 107 (1994).256. William W. Fisher III, Reconstructing the Fair Use Doctrine, 101 Harv. L. Rev.1659 (1988); Fitzgerald, supra note 109.257. A & M Records, Inc. v. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001)258. See, e.g., Associated Press v. United States, 326 U.S. 1, 20 (1945); TurnerBroadcasting Sys., Inc. v. FCC, 512 U.S. 622, 662-64 (1994) (“The basic tenet ofnational communication policy is that the widest possible dissemination of informationfrom diverse and antagonistic sources is essential to the welfare of the public.”).259. 17 U.S.C. § 1201(f)(4) (Supp. IV 1998).

engineer the industry standard software in an attempt to makesoftware that can be interoperable (conversant) with the industrystandard. To be able to successfully reverse engineer software, inmost instances, one must copy (and in some cases “borrow” partsof) the software, which is technically an infringement of thecopyright owner’s exclusive rights over reproduction.260In the United States, the courts have employed the fair usedoctrine to mediate this issue,261 while more recently, Congresshas legislated on the notion of interoperability in relation tocircumvention devices in the DMCA. Fair use defines theappropriate balance between a monopoly right given as anincentive for innovation and the public interest in the free flow ofinformation for a variety of cultural reasons.262In Australia the Copyright Law Review Committee (CLRC)263recommended adopting a more broad-based fair use right similarto the U.S. model in place of the existing narrower and morespecific fair dealing exceptions.264 However, the High Court ofAustralia, in the case of Data Access Corp. v. Powerflex ServicesPty. Ltd.,265 refused, in the absence of legislative direction, toendorse the legitimacy of reversing engineering in an instance ofliteral copying.266 The government has since enacted a (partial)legislative solution through amendment to the Copyright Act,1968 in 1999.267 The reproduction or adaptation of a computerprogram for the purpose of obtaining information necessary to theindependent creation of a new program or a device to connect toand interoperate with a program or any other program is, in260. See Anne Fitzgerald & Cristina Cifuentes, Interoperability and ComputerSoftware Protection in Australia, 4 Computer Telecomm. L. Rev. 271, 271 (1998).261. See generally Sega Enters. Ltd. v. Accolade, Inc., 977 F.2d 1510 (9th Cir. 1992)(holding that the use of copyrighted computer work to gain understanding ofunprotected functional elements was fair use of the copyrighted work).262. See Fitzgerald, supra note 109, at 153.263. Copyright Law Review Committee, supra note 60, at Part 1.264. See, e.g., Copyright Act, 1968, §§ 40- 42 (Austl.) (discussing fair dealing forpurposes of research or study, criticism or review, and reporting news).265. [1999] H.C.A. 49 (Austl.).266. Id.267. See Copyright Amendment (Computer Programs) Act, 1999, (Austl.); AnneFitzgerald & Cristina Cifuentes, Pegging Out the Boundaries of Computer SoftwareCopyright: The Computer Programs Act and the Digital Agenda Bill, in Going Digital2000: Legal Issues for E-Commerce Software and the Internet 37 (Anne Fitzgerald etal. eds., 2d ed. (2000)).

certain circumstances, exempted from infringement.268In European Union member countries, specific exceptionshave been enacted in national copyright laws,269 as required bythe 1991 EC Directive on the Legal Protection of ComputerPrograms.270 Article 6 of the Directive provides that copying andadaptation occurring in the decompilation of software code ispermitted where it is “indispensable to obtain the informationnecessary to achieve the interoperability of an independentlycreated computer program with other programs.”2712. Porting to Platforms of ChoiceTwo recent cases highlight a more specific and immediatechallenge for the principle of fair use for interoperability in thedigital environment. The challenge is to enable softwareapplications to be separated from and interoperable with a varietyof operating systems and associated hardware (platforms).272 It isin facilitating this choice with respect to operating platforms (e.g,open versus proprietary) that interoperability will allow the userto speak in their choice of differently coded/digital voices, therebyenhancing the goal of digital diversity.In the U.S. case of Sony Computer Entertainment, Inc. v.Connectix Corp.,273 the Ninth Circuit Court of Appeals held firmlyin favour of reverse engineering274 of software in order to allowdifferent software products to be ported or joined (i.e.,interoperate) with different hardware, firmware or softwareplatforms.275 In that case, Judge Canby, delivering the opinion ofthe court, explained that the intermediate276 copying that268. Copyright Act, 1968, § 47D (Austl.), amended by Copyright Amendment(Computer Programs) Act, 1999, (Austl.).269. See, e.g., Copyright Designs and Patents Act, 1988, §§ 50A, 50B, 50C (UK)(discussing acts done under statutory authority).270. Council Directive 91/250, 1991 O.J. (L 122) 42.271. Id.272. Platform is defined, in terms of computer science, as the basic technology of acomputer system’s hardware and software that defines how a computer is operated anddetermines what other kinds of software can be used, at http://www.dictionary.com/cgibin/dict.pl?term=platform(last visited Nov. 3, 2001). See also Microsoft, 84 F. Supp. 2dat 12 (stating that because the operating system supports applications while closelyinteracting with the PC system’s hardware, the operating system serves as a platform).273. 203 F.3d 596 (9th Cir. 2000).274. See id. at 599-601.275. See id. at 609.276. Id. at 598-99. Intermediate copying is that which is undertaken in the process

occurred was legitimate because it merely facilitated the copyingof the unprotected (non-copyright) function or idea of thesoftware.277 Interestingly, the existence of a patent would havemade things more difficult for Connectix.278 However, the courtclearly stated that intermediate copying is allowable where it isused to determine function that can facilitate the making oftransformative (better or extended) products.279Connectix had developed software that emulated the SonyPlayStation on a personal computer.280 In other words, theirVirtual Game Station allowed consumers to play Sony games ontheir personal computer and not just on the Sony PlayStationconsole. This allowed portability that the court also seemed tosupport in its discussion of transformative use: “The productcreates a new platform, the personal computer on whichconsumers can play games designed for Sony PlayStation. Thisinnovation affords opportunities for game play in newenvironments . . . .”281 The court was not concerned with thenumber of intermediate copies that had been made nor that theend product (even though containing a different source code, yetthe same function) competed with the Sony PlayStation.282In contrast, Reimerdes, which was brought pursuant to theDMCA for offering, providing or otherwise trafficking in a deviceof developing the final product which itself may not, as in the Connectix case, be a copy.277. Id. at 602-03. The judge remarked:[I]n the case of computer programs the idea/expression distinction posesdistinct ‘unique problems’ because computer programs are in essence‘utilitarian articles - articles that accomplish tasks. As such they containmany logical, structural and visual display elements that are dictated by thefunction to be performed, by considerations of efficiency, or by externalfactors such as compatibility requirements and industry demands’ . . . the fairuse doctrine preserves public access to the ideas and functional elementsembedded in copyrighted computer software programs.Id. at 603.278. See Sony, 203 F.3d at 596; Maureen A. O’Rourke, Toward a Doctrine of FairUse in Patent Law, 100 Colum. L. Rev. 1177 (2000); Richard H. Stern, Scope ofProtection Problems with Patents and Copyrights on Methods of Doing Business, 10Fordham Intell. Prop. Media & Ent. L.J. 105 (1999). It may be necessary in the not toodistant future to state the rationale more simply in terms of allowing copying on thebasis of increasing portability and diversity.279. See Sony, 203 F.3d at 602-03.280. Id. at 599.281. Id. at 606.282. Id. at 604 (this might be termed “horizontal” interoperability); see O’Rourke,supra note 281, at 1213, 1223.

for circumventing a technological protection measure (TPM), afifteen year old boy allegedly cracked the encryption code (CSS)283of the software lock (a technological protection measure) employed(it was argued) to prevent easy copying of the content of DigitalVersatile Discs (DVDs).284 Judge Kaplan, in his interim and finaljudgement,285 held that the reverse engineering exception of theDMCA286 could not be invoked. In this case, the TPM was CSSand the circumvention device was DeCSS.287 The defendant wasalleged in the final instance to be linking to websites that allowedthe downloading of DeCSS.288The defendants argued that reverse engineering of thesoftware lock was needed in order to play DVDs on other283. Reimerdes, 82 F. Supp. 2d at 214. The court explained that:DVDs contain motion pictures in digital form, which presents an enhanced risk ofunauthorized reproduction and distribution because digital copies made from DVDs donot degrade from generation to generation. Concerned about this risk, motion picturecompanies, including plaintiffs, insisted upon the development of an access control andcopy prevention system to inhibit the unauthorized reproduction and distribution ofmotion pictures before they released films in the DVD format. The means now in use,Content Scramble System or CSS, is an encryption-based security and authenticationsystem that requires the use of appropriately configured hardware such as a DVDplayer or a computer DVD drive to decrypt, unscramble and play back, but not copy,motion pictures on DVDs. CSS has been licensed to hundreds of DVD playermanufacturers and DVD content distributors in the United States and around theworld.Id.284. Id. at 214 (stating that “DVDs are five-inch wide discs that, in thisapplication, hold full-length motion pictures. They are the latest technology for privatehome viewing of recorded motion pictures. This technology drastically improves theclarity and overall quality of a motion picture shown on a television or computerscreen.”).285. See Reimerdes, 82 F. Supp. 2d at 217-18; Reimerdes, 111 F. Supp. 2d at 319-20.286. 17 U.S.C. § 1201(f) (1994 & Supp. IV 1998).287Reimerdes, 82 F. Supp. 2d at 216-17. The court explained that:CSS is a technological measure that effectively controls access to plaintiffs’ copyrightedmovies because it requires the application of information or a process, with theauthority of the copyright owner, to gain access to those works. Indeed, defendantsconceded in their memorandum that one cannot in the ordinary course gain access tothe copyrighted works on plaintiffs’ DVDs without a “player key” issued by the DVDCCA that permits unscrambling the contents of the disks. It is undisputed also thatDeCSS defeats CSS and decrypts copyrighted works without the authority of thecopyright owners. As there is no evidence of any commercially significant purpose ofDeCSS other than circumvention of CSS, defendants’ actions likely violated Section1201(a)(2)(B).Id.288. See id. at 214-15.

platforms such as Linux.289 They argued that DeCSS wasnecessary to achieve interoperability290 between computersrunning on the Linux system and DVDs; therefore, theinteroperabilty exception in the DMCA was enlivened. The judgeexplained that he could not accept such an argument for threereasons:First, defendants have offered no evidence to support thisassertion. Second, even assuming that DeCSS runsunder Linux, it concededly runs under Windows—a farmore widely used operating system—as well. It thereforecannot reasonably be said that DeCSS was developed “forthe sole purpose” of achieving interoperability betweenLinux and DVDs. Finally, and most important, thelegislative history makes it abundantly clear that Section1201(f) permits reverse engineering of copyrightedcomputer programs only and does not authorizecircumvention of technological systems that controlaccess to other copyrighted works, such as movies.291The defendants also made the argument that they were engagedin a fair use under Section 107 of the Copyright Act.292 The judgerejected this argument:Section 107 of the Act provides in critical part thatcertain uses of copyrighted works that otherwise would bewrongful are “not . . . infringement[s] of copyright.”Defendants, however, are not here sued for copyrightinfringement. They are sued for offering to the publicand providing technology primarily designed tocircumvent technological measures that control access tocopyrighted works and otherwise violating Section1201(a)(2) of the Act. If Congress had meant the fair usedefense to apply to such actions, it would have said so.293A crucial issue arising from such reasoning is whether the289. See id. at 218.290. See id. (This might be termed “vertical” interoperability); see O’Rourke, supranote 198, at 1213, 1219-1223.291. See Reimerdes, 82 F. Supp. 2d at 218.292. See id. at 219.293. Id.; see also Kamiel J. Koelman & Natali Helberger, Protection ofTechnological Measures, in Copyright and Electronic Commerce: Legal Aspects ofElectronic Copyright Management (P. Bernt Hugenholtz ed. 2000).

DMCA creates an exclusive right in the person or entity settingthe technological protection measure to control access to copyrightworks. That would allow a person or entity to protect raw data(embodied in a copyright work) by encasing it in a TPM or toprevent fair use of copyright material by encasing it in the TPM—some suggest this is akin to locking a book or a database in a roomor bank vault.294 Reimerdes, while acknowledging somelimitations, intimates that the DMCA does create a right tocontrol access to copyright works and that this does not violate ofthe First Amendment, as this incidental impact on protectedexpression is acceptable in light of the overall objective ofCongress to protect copyright in the digital environment.295 Itmight be suggested that such an approach serves to eliminate fairuse, create rights in unprotected data and portrays overly broadregulation.296 An even more pressing point to appreciate is thatfair use rights to reverse engineer software architecture in thename of interoperability, where that software is a technologicalprotection measure or lock, are severely restricted by the DMCAas interpreted in this case.One further point arising in the case and throwing light onthe meaning of the DMCA must be noted. In the process ofarguing fair use, the defendants raised the Sony defence. Thejudge responded:Defendants claim also that the possibility that DeCSSmight be used for the purpose of gaining access tocopyrighted works in order to make fair use of thoseworks saves them under Sony Corp. v. Universal CityStudios, Inc. But they are mistaken. Sony does not applyto the activities with which defendants here are charged.Even if it did, it would not govern here. Sony involved aconstruction of the Copyright Act that has been overruled294. See Pamela Samuelson, Intellectual Property and the Digital Economy: Whythe Anti-Circumvention Regulations Need to be Revised, 14 Berkeley Tech. L.J. 519,538-43 (1999).295. See Reimerdes, 111 F. Supp. 2d at 322-23, 330; see also Council Directive01/29/EC of 22 May 2001 on the Harmonisation of Certain Aspects of Copyright andRelated Rights in the Information Society, 2001 O.J. (L 167) 10, art. 6(1) (prohibitingactual circumvention); cf. Copyright Amendment (Digital Agenda) Act, 2000, (Austl.)(failing to expressly prohibit actual circumvention).296. In considering whether such legislation should be allowed to protect noncopyrightdata see Benkler, supra note 10; Patry, supra note 128; Reichman &Samuelson, supra note 22.

y the later enactment of the DMCA to the extent of anyinconsistency between Sony and the new statute. Sonywas a suit for contributory infringement brought againstmanufacturers of videocassette recorders on the theorythat the manufacturers were contributing to infringinghome taping of copyrighted television broadcasts. TheSupreme Court held that the manufacturers were notliable in view of the substantial numbers of copyrightholders who either had authorized or did not object tosuch taping by viewers. But Sony has no applicationhere.When Sony was decided, the only question was whetherthe manufacturers could be held liable for infringementby those who purchased equipment from them incircumstances in which there were many noninfringinguses for their equipment. But that is not the questionnow before this Court. The question here is whether thepossibility of noninfringing fair use by someone who gainsaccess to a protected copyrighted work through acircumvention technology distributed by the defendantssaves the defendants from liability under Section 1201.But nothing in Section 1201 so suggests. By prohibitingthe provision of circumvention technology, the DMCAfundamentally altered the landscape. A given device orpiece of technology might have “a substantialnoninfringing use, and hence be immune from attackunder Sony’s construction of the Copyright Act—butnonetheless still be subject to suppression under Section1201.” Indeed, Congress explicitly noted that Section1201 does not incorporate Sony.. . . The fact that Congress elected to leave technologicallyunsophisticated persons who wish to make fair use ofencrypted copyrighted works without the technical meansof doing so is a matter for Congress unless Congress’decision contravenes the Constitution, a matter to whichthe Court turns below.297In summary, the DMCA provides an exception for297. Reimerdes, 111 F. Supp. 2d at 323-24.

circumventing a TPM in order to look at a program to identify andanalyse those elements necessary to achieve interoperability inthe name of reverse engineering, but only to the extent permittedby copyright law, most commonly where a right to reverseengineer under the fair use doctrine is available.298 Thus, thevalue of interoperability and digital diversity was seen inReimerdes as being of less importance than the copyright ownersexclusive right to reproduction or to control access. This was ahard case and this type of fact scenario needs much more thought.However, to prevent the porting of informational products todifferent platforms stifles diversity and allows one entity tocontrol the structure or architecture for disseminating informationand constructing meaning. Is this acceptable? Obviously a verybroad right to reverse engineer given as an exception to copyright(or circumvention?) law reduces the incentive for developingsoftware or content. However, this must be balanced against theneed for interoperability and digital diversity in softwarearchitecture.2993. Fair Use: NapsterFair use was also unsuccessfully argued in the recent Napsterdecision, although many would argue that this is a case in whichthe social benefits of the user outweigh the economic arguments ofcopyright protection.300There have been a number of decisions concerning what isknown as “mp3.” Mp3 is a format/software that allows digitizedmusic to be compressed and digitally distributed without losingsound quality.301 The recording industry throughout the world298. See, e.g., RealNetworks, Inc. v. Streambox, Inc., 2000 WL 127311 (W.D. Wash.2000); Sony Computer Entm’t of Am., Inc. v. Gamemasters, 87 F. Supp. 2d 976 (N.D.Cal. 1999).299. See Lessig, supra note 42.300. See Fred von Lohmann, Peer-to-Peer File Sharing and Copyright Law AfterNapster, Cal. Law., Jan. 2001, at 42, available athttp://www.eff.org/IP/P2P/Napster/20010309_p2p_exec_sum.html (last visited Nov. 17,2001).301. See Recording Indus. Ass’n of Am. v. Diamond Multimedia Sys., Inc., 180 F.3d1072, 1073-74 (9th Cir. 1999).Until recently, the Internet was of little use for the distribution of musicbecause the average music computer file was simply too big: the digitalinformation on a single compact disc of music required hundreds of computerfloppy discs to store, and downloading even a single song from the Internettook hours. However, various compression algorithms (which make an audio

has been greatly concerned with the proliferation of mp3formatted music files as this has (they say) the potential toundermine the whole system of copyright in popular music. Thetechnology is such that a song can be “ripped” off a CD, uploadedto the Internet for copying, distribution, consumption andpotentially downloading or “burning” onto a compact disc.In many instances, the act of copying is obvious and arguablya breach of copyright. However, lawsuits so far have not beenagainst the end consumer for direct copyright infringement butagainst the machine or technology manufacturers or developersfor indirectly breaching copyright through facilitatinginfringement. Napster involved suits by the recording industryagainst the company that developed and distributed software thatenabled people to reference and access mp3 files available via theInternet. A “live” index created and centrally stored at Napsterallowed online users to share mp3 files/songs. For instance, if Iwant to listen to a song called “Red River,” I would use myNapster software to access a list of people currently online fromthe Napster central server who are registered as having a copy ofthe song; then, I would request the file directly from the personwho has it stored on their hard drive. The file is not sent to me byNapster, but by the party who is registered as having a copy. Thisis sometimes referred to as peer-to-peer file sharing.302 Theimportant point is that Napster merely facilitates the delivery ofthe copied song, it does not do any copying itself. Should they beliable for copyright infringement? Does the digital property of thecopyright owner control this situation?In 1984, the United States Supreme Court, analogizing thefile “smaller” by limiting the audio bandwidth) now allow digital audio files tobe transferred more quickly and stored more efficiently. MPEG-1 AudioLayer 3 (commonly known as “MP3”) is the most popular digital audiocompression algorithm in use on the Internet, and the compression itprovides makes an audio file “smaller “ by a factor of twelve to one withoutsignificantly reducing sound quality. MP3’s popularity is due in large part tothe fact that it is a standard, non-proprietary compression algorithm freelyavailable for use by anyone, unlike various proprietary (and copyright-secure)competitor algorithms. Coupled with the use of cable modems, compressionalgorithms like MP3 may soon allow an hour of music to be downloaded fromthe Internet to a personal computer in just a few minutes.Id.302. See Andrew Oram, Peer to Peer: Harnessing the Power of DisruptiveTechnologies (2001).

“staple article of commerce” doctrine in U.S. patent law,303 heldthat Sony was not liable for copyright infringement for making avideo recorder because a video recorder has substantial noncopyrightinfringing uses, such as time shifting, i.e., recordingmaterial so that you can watch it at a later date.304 CouldNapster rely on the same Sony defence?Ultimately, Napster was sued for copyright infringement onthe basis of contributory or vicarious copyright infringement.305Before those actions could be established, the court had todetermine whether there had been any direct infringement by theend consumers.306 The court concluded that there had been.307The court rejected arguments based on fair use sampling(listening to see if they liked the music) and space shifting(location of where one listened to the music).308 The court rejectedarguments based on sampling, saying that there was evidence toshow that sampling ruined the full commercial exploitation of thesongs, and on space shifting, saying this was not like Sony in thatspace shifting in this instance opened up copied product to theworld.309The next issue the court considered was whether there wascontributory infringement. They listed the following elements:Did Napster know or should they have known of theinfringement—actual or constructive knowledge? Yes,they knew users were swapping infringing files;310Did Napster materially contribute to the infringement?Yes, they provided software;311Were there any defences? No. The court assessed theSony substantial non-infringing use defence, explainingthat current and future uses need to be considered, butdecided the Sony defence did not apply once knowledge303. 35 U.S.C. § 271(c) (1994 & Supp. V 1999).304. Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417 (1984).305. A & M Records, Inc. v. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001).306. Id. at 1013.307. Id. at 1016.308. Id. at 1018-19.309. Id. at 1019.310. See id. at 1020-22.311. See Napster, 239 F.3d at 1022.

was evident;312Did they sell the technology knowing people would use itfor copyright infringement? In the Sony case the courtsaid no—in Napster, the court said yes.313The next issue was that of vicarious infringement. The courtconsidered the following elements:Did Napster have authority to control or superviseinfringement? Yes, because it could monitor songnames;314Did Napster gain a direct financial benefit frominfringement? Yes, because it gained more users andvalue through the infringement.315The court also stated that the Sony defence does not apply toan action for vicarious copyright liability.316 Waiver, impliedlicence and copyright misuse arguments were all rejected.317 Thescope of the injunction was readjusted so that song names had tobe delivered to Napster by the record companies.318 The court ofappeals left open the question of the application of the ISPimmunity provisions of the DMCA to Napster.319 Napster is nowengaged in a process of signing licence agreements with musiccompanies in the increasingly pay-per-view/hear world.The key question arising out of this case is whether copyrightis killing technological innovation. Should the developer orinnovator be liable for copyright infringement in this case? Whatwill this do for technological innovation? This is a case ofcopyright versus cultural/social/communicative experiment—should copyright give way a little?Consider Sony, where it was held that a “finding ofcontributory infringement effectively extends the grant of power toencompass the accused instrumentality—where a technology hasother lawful uses such a rule would block the wheels of312. See id. at 1020.313. Id. at 1020-21.314. Id. at 1023.315. Id.316. Id. at 1022.317. See Napster, 239 F.3d at 1026-27.318. Id. at 1027.319. Id. at 1025.

commerce.”320 Is innovation stifled by this decision or was thissimply a clear case of copying facilitated by a knowingintermediary? Should the digital property owner be able tocontrol innovation in technology that may represent a risk to thevalue of that property? So far, in the context of the new digitaltechnologies, the answer from the courts is likely to be “yes.”In light of the foregoing discussion it can be seen that fair useis an extremely important variable in the definition of digitalproperty and will be raised in many instances, with varyingdegrees of success.C. Privacy LawThe notion of privacy means at very least the right to controlpersonal information/informational self-determination.321 Thisright has long been recognised in international law.322 Privacy ordata protection law promises to be a key determinant ofinformational property in the digital age.While use of private information by government and publicbodies has been regulated in the United States, Canada andAustralia during the last twenty years, the flow of data in theprivate sector has been subject to very little legal constraint. In1995, the European Union, in response to concerns over theinvasive nature of the Internet and its tremendous capacity totrace and profile individual identity, promulgated the EuropeanDirective on Data Protection.323 In short, the Directive requiresmember states to enact legislation covering the processing of datacollection in the private sector.324 The purpose for which theinformation was gathered must be disclosed at the point of receipt,320. Sony, 464 U.S. at 441.321. See Lee A. Bygrave & Kamiel J. Koelman, Privacy, Data Protection andCopyright: Their Interaction in the Context of Electronic Copyright ManagementSystems, in Copyright and Electronic Commerce: Legal Aspects of Electronic CopyrightManagement 59, 64-65 (P. Bernt Hugenholtz ed., 2000); see also Moore v. Regents ofthe University of California, 793 P.2d 479 (Cal. 1990); Restatement (Second) of Torts §652 (1977); Julie E. Cohen, Examined Lives: Informational Privacy and the Subject asObject, 52 Stan. L. Rev. 1373 (2000); Charles Fried, Privacy, 77 Yale L.J. 475 (1968);William L. Prosser, Privacy, 48 Cal. L. Rev. 383 (1960); Eugene Volokh, Freedom ofSpeech and Information Privacy: The Troubling Implications of a Right to Stop PeopleFrom Speaking About You, 52 Stan. L. Rev. 1049 (2000); Samuel D. Warren & Louis D.Brandies, The Right to Privacy, 4 Harv. L. Rev. 193 (1890).322. Universal Declaration of Human Rights, supra note 15, at art. 12.323. Council Directive 1/95, 1995 O.J. (C 93) 1.324. Id.

and the information may only be used for that purpose.325Individuals have the right to see data on themselves, yet full andinformed consent by the data subject may be used to overridemany of the obligations imposed by the Directive.326 Mostimportantly, Article 25 of the EU Directive stipulates that EUbusinesses cannot disclose data to members of third party statesunless it is shown that effective data protection regimes are inplace in those states. In April 2000 Canada responded by enactingsimilar obligations in the Privacy and Electronic DocumentsAct.327 Likewise, in December 2000 Australia enacted the PrivacyAmendment (Private Sector) Act.328 In the United States,legislation has been avoided. In its place, the United StatesDepartment of Commerce has promulgated safe harbour rules,which act as an optional self-regulatory structure. The efficacy ofsuch a regime is still in question.In the United States, state constitutional law and other formsof specific state privacy legislation329 may also be relevant in somecases; e.g., Californian Constitution Article 1, Section (1): “Allpeople are by nature free and independent and have inalienablerights. Among these are enjoying and defending life and liberty,acquiring, possessing, and protecting property, and pursuing andobtaining safety, happiness, and privacy.”330 Through thisprovision, “California accords privacy the constitutional status ofan inalienable right on a par with defending life and possessingproperty.”331 “Privacy is protected not merely against stateaction; it is considered an inalienable right that may not beviolated by anyone.”332More recently, litigation in the United States against the325. Id.326. See id.; see also Charter of Fundamental Rights of the European Union art. 8(reinforcing this idea).327. Personal Information Protection and Electronic Documents Act, S.C., ch. 5(2000) (Can.).328. Privacy Amendment (Private Sector) Act, 2000, no. 155 (Austl.).329. See, e.g., Cal. Civ. Code §§ 1749.60 (Supermarket Club Card Legislation)(2001), 1708.7 (2001) (Cyberstalking); Dorothy J. Glancy, Privacy and IntelligentTransportation Technology, 11 Santa Clara Computer & High Tech. L.J. 151, 189-202(1995).330. Cal. Const. art. 1, § 1.331. Luck v. Southern Pac. Transp. Co., 267 Cal. Rptr. 618, 625 (Cal. Ct. App.1990) (citations omitted), cert. denied, 498 U.S. 939 (1990).332. Porten v. Univ. of S.F., 134 Cal. Rptr. 839, 842 (Cal. Ct. App. 1976) (citationsomitted).

anner advertisement company DoubleClick addressed thelegality of “cookies” (technology that allows the viewing habits ofyour computer to be traced).333 The legal arguments madepursuant to federal law, which were based on the ElectronicCommunications Privacy Act, Wire Tapping Act, and theComputer Fraud and Abuse Act, against DoubleClick weredismissed.334 The more interesting state law claims concerningthe privacy tort,335 unjust enrichment and possibly trepass tochattels were not dealt with by the federal court. It remains to beseen whether the privacy tort (intrusion upon seclusion) cansuccessfully be argued in a case involving cookie technology. Andwhen that happens, courts will have to ask: Would it be illegal forsomeone to follow you around in real space and note your buyinghabits?336 Have you put yourself in public view by logging ontothe network? There may also be a First Amendment challengeagainst such state based privacy tort, unjust enrichment andtrespass actions.337Increasingly, privacy regimes for the private sector willregulate the way in which private data can be exploited in themarket place. The message to be taken from the EuropeanDirective is that you should not collect private information unlessyou need to, but if you must then explain the purpose, collect it ina fair way, allow access to it by the data subject and gain theirconsent for using it in secondary ways. Surreptitious or covertways of collecting information should not be encouraged; howeverif the information can be collected without identifying the subjectthen the law will usually not be breached.338 Although, in those333. In re DoubleClick Inc. Privacy Litig., 154 F. Supp. 2d 497 (S.D.N.Y. 2001).334. See id.335. See Restatement (Second) of Torts § 652 (1997).336. See generally Nader v. General Motors Corp., 255 N.E.2d 765 (N.Y. 1970)(holding that mere information gathering about a particular person does not give riseto action for invasion of privacy where information is open to the public and voluntarilygiven).337. See Volokh, supra note 329; U.S. West, Inc. v. FCC, 182 F.3d 1224 (10th Cir.1999); see generally Jennifer Barrett et al., Data Privacy Laws and the FirstAmendment: A Conflict, 11 Fordham Intell. Prop. Media & Ent. L.J. 1 (2000)(discussing First Amendment and data protection issues); Solveig Singleton, PrivacyVersus the First Amendment: A Skeptical Approach, 11 Fordham Intell. Prop. Media &Ent. L.J. 97 (2000) (discussing same).338. See, e.g., Council Directive 95/46, art. 2(a), 1995 O.J. (L 281) (noting “‘personaldata’ shall mean any information relating to an identified or unidentifiable naturalperson (‘data subject’); an identifiable person is one who can be identified, directly orindirectly, in particular by reference to an identification number or to one or more

cases where the manner of collection allows fairly preciseidentification, even without a name, such collection will not beallowed.An interesting question that arose in the DoubleClick caseand that is allied to the notion of database rights is whether thedata subject has a right to claim payment for the data that hasbeen collected and is being exploited. There, the court explained:Plaintiffs also contend that they have suffered economicdamages consisting of the value of: (1) the opportunity topresent plaintiffs with advertising; and (2) thedemographic information DoubleClick has collected.Essentially, they argue that because companies payDoubleClick for plaintiffs’ attention (to advertisements)and demographic information, the value of these servicesmust, in some part, have rightfully belonged to plaintiffs.They point to AOL in which the court appeared to holdthat damage to “reputation and goodwill” countedtowards the damage threshold and argue that, by thesame logic, the economic value of their attention anddemographic information should count as well.Even assuming that the economic value of plaintiffs’attention and demographic information could be countedtowards the monetary threshold—a dubiousassumption—it would still be insufficient. We do notcommonly believe that the economic value of ourattention is unjustly taken from us when we choose towatch a television show or read a newspaper withadvertisements and we are unaware of any statute orcaselaw that holds it is. We see no reason why Web siteadvertising should be treated any differently. A personwho chooses to visit a Web page and is confronted by atargeted advertisement is no more deprived of hisattention’s economic value than are his off-line peers.Similarly, although demographic information is valuedhighly (as DoubleClick undoubtedly believed when it paidover one billion dollars for Abacus), the value of itscollection has never been considered a economic loss tofactors specific to his physical, physiological, mental, economic, cultural or socialidentity”).

the subject. Demographic information is constantlycollected on all consumers by marketers, mail-ordercatalogues and retailers. However, we are unaware ofany court that has held the value of this collectedinformation constitutes damage to consumers or unjustenrichment to collectors.339While the court was apprehensive about recognising this personalclaim for the value of the private data, some academics have goneas far as suggesting that people should be recognised in law aspossessing a property (intellectual property?) right in theirpersonal information.340 This would allow people to protect suchinformation through law and to charge for it in the market place.Samuelson has cautioned against such an approach, arguing thatit would distort the rationale for having property rights ininformation, which could lead to higher transaction costs and aworsened situation for the data subject.341 She prefers a form oflicensing that licenses private information along the same lines astrade secret licensing, where one contractually structures theusage rights of the information before disclosure.342It can be seen from this brief analysis that the growing law onprivacy or data protection, especially in the private sector, willhave a significant impact upon the shape and content of digitalproperty rights.IV. CONTRACTUALLY DEFINED DIGITAL PROPERTYThe contract, an agreement between two or more parties, isthe foundation of many privately ordered relationships and isincreasingly being used to regulate informational propertyentitlements in the digital environment.A. Contractually Created Informational Property RightsContractual rights can be used to extend public or legislativerights to informational value, especially in the area of data339. In re DoubleClick, 154 F. Supp. 2d at 525 (footnotes and citations omitted); seealso Weld v. CVS Pharmacy Inc., 10 Mass. L. Rptr. 217 (Mass. Super. 1999).340. See Cohen, supra note 321; see also Jessica Litman, InformationPrivacy/Information Property, 52 Stan. L. Rev. 1283 (2000).341. See Pamela Samuelson, Privacy as Intellectual Property, 52 Stan. L. Rev. 1125(2000).342. See id.

capital.343 While there may be no recognised legislativeintellectual property right in data, a contract can be used toregulate the way in which people use information that has beencollected to create contractual or privately ordered informationalproperty rights. The limit of a contract (besides some complexconstitutional issues) is that it is only enforceable against theparties to the contract and not the world at large. For instance, ifI buy a licence to use data that contains restrictions on my furthercopying, selling or exploiting the data, that licence will not bind astranger who copied the data.B. Contractual LicensingContract is also very important to the broader issue ofinformation licensing.344 In the information economyinformational products are usually licensed and not sold; thelicense is the product. In most instances you do not receive anownership right to anything but rather a right to use information.In real space, when I buy a book I do not obtain ownership of thecopyright owner’s right to reproduce the book, but I do gainownership of the physical thing called the book. Withinformational products, such as software games and digitalimages, you are usually only given a licence to use the informationfor specific purposes, and to this end, you must read the licenceclosely to determine your user rights. In essence, the contractuallicence determines “exploitation” and “user” rights in informationand in this way contract is used as a mechanism for definingdigital property.C. Formation of Contractual LicencesAn issue vitally important to this contractual definition ofdigital property is the way in which contractual licences areformed. In order to determine informational rights, it is343. See ProCD, Inc. v. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996).344. See Gail E. Evans & Brian F. Fitzgerald, Information Transactions UnderUCC Article 2B: The Ascendancy Of Freedom of Contract in the Digital Millennium, 21U. New S. Wales L.J. 404 (1998), available at http:// www.law.unsw.edu.au/publications/journals/unswlj/ecommerce/evans.html;Robert W. Gomulkiewicz, The LicenseIs The Product: Comments on the Promise of Article 2B for Software and InformationLicensing, 13 Berkeley Tech. L.J. 891 (1998); David Nimmer et al., The Metamorphosisof Contract into Expand, 87 Cal. L. Rev. 17 (1999). Licenses are contracts that defineuser rights in relation to information products.

important to know whether a contract is in existence and uponwhat terms. This question has been raised in the context ofshrink-wrap and click-wrap/click-through contracts. “Shrinkwrapcontracts” are normally used where the product, e.g.,software, is sold in a store. The terms of the contract are shrinkwrappedaround the product in fine print and in most instancesnot fully disclosed until after opening the package. “Click-wrap orclick-through contracts” are normally employed in an onlineenvironment, such as the Internet, and may, at best, involve theacceptance of the terms of the contract by clicking on a digital icondenoting consent after having been shown the terms. In someclick wrap cases, which are highly questionable from a legalperspective, terms are not clearly displayed prior to the buyerentering into the agreement. In the recent decision of Specht v.Netscape Communications Corp and AOL,345 a third type ofsituation called “browse-wrap” was outlined. A browse-wrapscenario arises where the Internet consumer is not required tonegotiate an “I agree” button or icon or view the licence termsbefore gaining access to the product. Instead, the terms arelocated behind a button/icon on the website that the consumer isnot forced to view or engage. In Specht, the court held that thebrowse-wrap scenario would not usually give rise to a validcontract.346In ProCD, Inc. v. Zeidenberg,347 the court held that shrinkwrap,and arguably click-wrap, licences are enforceable in certaincircumstances.348 ProCD developed a product known as SelectPhone (the compilation of over 3,000 telephone directories on CD),which was sold in stores for $150.349 ProCD invested over $10million in developing this product.350 Mindful of the Feist decisionand the inability of copyright law to protect raw data, ProCDsought to commodify informational value through a contractual345. 150 F. Supp. 2d 585 (S.D.N.Y. 2001).346. See id. at 594.347. 86 F.3d 1447 (7th Cir. 1996).348. Id. at 1449; see, e.g., Register.com v. Verio Inc., 126 F. Supp. 2d 238 (S.D.N.Y.2001); Hotmail Corp. v. Van Money Pie Inc., 47 U.S.P.Q. 2d. 1020 (N.D. Cal. 1998);Lemley et al., Software and Internet Law 494-95 (2000); see also Hill v. Gateway 2000,Inc., 105 F.3d 1147 (7th Cir. 1997), cert. denied, 118 S. Ct. 47 (1997); Nat’l BasketballAss’n v. Motorola, Inc., 105 F.3d 841 (2d Cir. 1997).349. ProCD, 86 F.3d at 1449.350. Id.

agreement.351 Shrink-wrapped around the CD in plastic orcellophane was a detailed user agreement or licence, whichcontained a term stipulating that the application program orinformation should not be on sold for commercial purposes.352Zeidenberg allegedly purchased the CD, went home and loaded itonto a website where it was sold at a price less than that chargedby ProCD.353 He was sued for breach of contract and argued thatnothing in the contract prevented him from selling theinformation.354 He argued that, at the point of purchase, it wasimpossible to read all the terms of the shrink-wrapped licence;therefore, he had no adequate notice of those terms and was notbound by them.355 The court held that the economics of the newinformation economy suggested that he should be bound by theshrink-wrapped terms and furthermore, if he did not like them, hecould have taken the CD back, but by acquiescing and choosing tokeep and use the informational product he was bound by theterms.356 While many have questioned whether this reasoning isfair, it is now embodied in sections 208 and 209 of UCITA, underthe umbrella of layered contracting.357Judge Easterbrook rejected the argument that the contractuallicence was preempted by section 301(a) of the U.S. Copyright Act.The argument was that creating copyright-like rights in data, thecontractual user agreement was in contravention of section 301 ofthe Copyright Act, which preempts states from legislating on theissue of copyright.358 Judge Easterbrook held that contractual351. Id.352. Id.353. Id. at 1450.354. Id. at 1450-52.355. ProCD, 86 F.3d at 1450-51.356. See id.357. UCITA is a model law promulgated for adoption by the United States thatcreates a “sale of goods” styled regime for the licensing of software and otherinformational transactions. It is argued that the process of transacting software andother informational products is not adequately covered by existing sale of goods typelegislation, which finds it hard to classify software. In the case law software issometimes classified as a good and sometimes as a service; leading commentators tolabel software the digital chameleon. UCITA aims to avoid this debate by creating asui generis regime governing the formation, performance and termination ofinformation transactions. It has been highly controversial in its content and has onlybeen adopted by two U.S. states: Maryland and Virginia. Uniform ComputerInformation Transactions Act, available athttp://www.law.upenn.edu/bll/ulc/ucita/ucita01.htm.358. See ProCD, 86 F.3d at 1453.

information rights and legislated copyrights were fundamentallydifferent.359 He explained that contract is a mechanism forprivate ordering that applies only between the parties to thecontract whereas the Copyright Act is public legislation thatapplies to the world at large.360Some critics have said that non-negotiated mass marketlicences are much more like legislation than contract and shouldbe closely scrutinised in terms of preemption.361 Lemley suggeststhat the weight of the ProCD decision should not be overstated asit goes against the majority of judicial opinion on the issue.362 InAustralia, the courts have yet to explicitly determine the legalvalidity of shrink-wrap and click-wrap contracts. Such contractswill inevitably be challenged, as in the United States, on the basisthat they do not adequately disclose the terms of the contract priorto the entering of the agreement.363 The issue of the formation ofcontractual licences will continue to be important to thecomplexion and definition of digital property because knowingwhether a contract exists, and upon what terms, is intimatelyconnected to informational exploitation and user rights.D. Contractual Allocation of Domain NamesContract has been even more prominent in the area of domainname allocation. Allocation of the premium Internet commercialtrading domain “.com” is now performed through a contractualregime created under the auspices of the Internet Corporation for359. Id. at 1454.360. Id.361. See Dennis S. Karjala, Federal Preemption of Shrinkwrap and On-LineLicenses, 22 U. Dayton L. Rev. 511 (1997); Mark A. Lemley, Beyond Pre-emption: TheLaw and Policy of Intellectual Property Licensing, 87 Cal. L. Rev. 111 (1999); MaureenA. O’Rourke, Copyright Preemption After the ProCD Case: A Market-Based Approach,12 Berkeley Tech. L.J. 53 (1997); David A. Rice, Public Goods, Private Contract, andPublic Policy: Federal Preemption of Software License Prohibitions Against ReverseEngineering, 53 U. Pitt. L. Rev. 543 (1992).362. See Mark A. Lemley, Software and Internet Law 490-93 (2000).363. See Mark Sneddon, Legal Liability and E-Transactions 33-34 (2000), availableat http://www.noie.gov.au/publications/NOIE/NEAC/publication_utz1508.pdf (lastvisited Nov. 16, 2001); see generally Thornton v. Shoe Lane Parking Ltd., 2 Q.B. 163(1970) (holding in a case regarding contract printed on parking lot ticket that sinceplaintiff did not know of the exemption condition and the defendants had not donewhat was reasonably sufficient to bring it to his notice, it did not except them fromliability); Balmain New Ferry Co. v. Robertson (1906) 4 C.L.R. 379 (Austl.) (holdingthat notice posted above wharf entrance was sufficient to show implied agreement toterms).

Assigned Names and Numbers (ICANN)364 requiringarbitration.365The recent string of personality cases concerningJuliaRoberts.com, Sting.com and Madonna.com are interestingexamples.366 Enterprising business people moved in early andregistered those domains. The famous stars behind the nameswanted them and sought to invoke the ICANN dispute resolutionpolicy, which involves compulsory arbitration. At the time ofregistration, the domain name registrants contractually bindthemselves, in the instance of a dispute, to enter arbitration andabide by the decision. If the registrant is found to have registeredand used the domain name in bad faith, vis-a-vis a trademark,they stand to lose that registration. Julia Roberts and Madonnawere successful in showing the necessary elements, including thatthey possessed a trademark, while Sting, due to the genericnature of the word and other difficulties, did not possess therequisite trademark.This system is in line for an overhaul as the legitimacy,consistency and accountability of the process is in question.367Nevertheless, let me make some observations. The ICANNsystem of resolving domain name disputes, which has dealt within excess of 3,000 cases since the beginning of the year 2000, hassought to protect informational value through contractuallyagreed principles. To date, the majority of the decisions havefavoured the traditional trademark holder and not the domainname speculator. Domain name speculation has been treated likea case of the theft of the investment value of the trademark.368The speculator has rarely been regarded as deserving of reward.Perhaps there should be arguments for value on both sides, but364. ICANN is available at http://www.icann.org.365. ICANN: Uniform Domain Names Dispute Resolution Policy (UDRP), art. 4, athttp://www.icann.org/udrp/udrp-policy-24oct99.htm (last visited Nov. 10, 2001).366. These decisions are available athttp://arbiter.wipo.int/domains/decisions/index.html.367. See generally A. Michael Froomkin, Wrong Turn in Cyberspace: Using ICANNto Route Around the APA and the Constitution, 50 Duke L.J. 17 (2000) (arguing thatthe use of ICANN to regulate in the stead of an executive agency violates fundamentalvalues and policies designed to ensure democratic control over the use of governmentpower, and sets a precedent that risks being expanded into other regulatory activities).368. See Brian F. Fitzgerald & Emma Sheehan, Trademark Dilution and theCommodification of Information: Understanding the ‘Cultural Command’, 3 MacarthurL. Rev. 61 (1999) (Austl.).

this value is difficult to apportion. For good or bad, what thissystem shows very clearly is a contractual regime being used toensure informational value is realised as wealth.Recently, Parisi v Net Learning Inc.,369 held that UniformDomain Names Dispute Resolution Policy (UDRP) determinationsare not arbitrations for the purposes of the Federal ArbitrationAct and thereby do not prevent access or review in the courts,explaining that “[a]lthough ICANN exerts quasi-governmentalsway over the growth and administration of the Internet, theUDRP is enforced through contract rather than regulation.”370To ease the pressure on the .com domain, to add greaterspecificity and to reduce disputes over domain names, ICANN hasapproved the allocation of seven new generic top level domains(TLDs): .aero, .biz, .coop, .info, .museum, .name, and .pro.371 Theinteresting question is how these domains will be contractuallyallocated. Will trademark holders be given priority and, if so,which trademark holders will gain priority? Will existing .comregistrants have the ability to block others registering in thesedomains?The first to be allocated will be .biz and .info. According tothe ICANN press release, .info will be an unrestricted TLD, opento any business or person to register for any purpose, while .bizwill be a restricted TLD, open only for commercial or businesspurposes.372 In November 2000, registry operators NeuLevel(.biz) and Afilias (.info) were among seven registry operators fornew TLDs selected by the ICANN Board of Directors following anopen process for submission of proposals. Agreements with theoperators of .aero, .coop, .museum, .name, and .pro are in theprocess of being completed.To provide for a stable, fair, and orderly introduction of thenew TLDs, NeuLevel and Afilias will employ different three-stepapproaches to launch their registry services:NeuLevel (.biz):1. Intellectual Property Claim Service. Beginning inlate May through the beginning of July, .biz will offer a369. 139 F. Supp. 2d 745 (E.D. Va. 2001).370. Id. at 747.371. See Announcements, at http://www.icann.org/announcements/icannpr15may01.htms(last visited Nov. 10, 2001).372. Id.

fee-based service for businesses to submit trademarkclaims before the application process is opened. Underthis service, domain-name applicants will be notified of atrademark owner’s claim when they submit anapplication for a claimed name and, if they decide to goahead with the registration, the business submitting theclaim will be notified of the registration and will be givena window of opportunity to challenge the registrationusing an administrative dispute resolution procedure.2. Domain Name Application and Selection. Beginningin July through late September, .biz will open theapplication process for businesses. Businesses will submittheir applications through registrars; these applicationswill be randomly selected in batches and awarded.Businesses can submit an unlimited number ofapplications but there will be a $2.00 fee for eachapplication for a .biz name.3. .biz Begins Operations. In October, .biz names thathave been awarded will become operational. .biz will usea Restriction Dispute Resolution Process (RDRP) similarto the ICANN UDRP process to resolve challenges relatedto the business-only restricted nature of the registry. 373Afilias (.info):1. Sunrise Period. Beginning in late June and endingin late July, Afilias will offer a 30 day “sunrise” periodduring which anyone with a registered nationaltrademark can apply through a registrar for a .infodomain name corresponding to the trademark.Trademark data will be provided at the time ofapplication and the term of the domain name registrationwill be for at least 5 years. To assure fair and equalaccess during the sunrise period, registrars will firstcollect applications and deliver them to Afilias. Afiliaswill then implement a multi-round batch processingsystem in which it will randomly select applications fromeach registrar’s batch. The World Intellectual Property373. Id.

Association (WIPO) will administer a challenge procedurethrough which any third party may challenge thesesunrise period registrations to ensure that they complywith the applicable requirements. This challenge processwill be available for approximately 120 days after Afiliasbegins accepting applications for .info domain names fromthe general public, ending in November.2. Domain Name Application and Registration. Within15 days after the conclusion of the sunrise period, Afiliaswill begin accepting applications for .info domain namesfrom the general public. To ensure that ordinary Internetusers have an equal chance to register names during theinitial launch phase, .info names will be registered usinga multi-round batch process similar to the one used forsunrise period applications. Using this process, Afiliaswill randomly select applications from batches submittedby each registrar. This process is an alternative to a purefirst-come-first-served system which could operate to theunfair advantage of those with mechanized registrationscripts. This batch process will last approximately threeweeks, after which Afilias will process applications on areal-time basis.3. .info Begins Operations. Names awarded during thesunrise period will become operational one week after thebeginning of the general application and registrationprocess. The remainder of the awarded info names willbecome operational promptly after Afilias processesthem.374These policies are contractual in nature and already there hasbeen concern over their implementation; NeuLevel was reportedlythreatened with legal action over the random allocation of .biz inthe face of competing trademark holders. Such a challenge willcall into question the power of this new contractually basedinformational property regime to effectively operate. In this typeof challenge, the power of contract to allocate “some” of the valueof digital property created under statute will be closely considered.This may, in turn, lead parties to further contemplate the values374. Id.

or policies upon which the contractual regime is founded.Overall, let me make this simple point. Contract is a vitalstrategy in creating and defining digital property.V. TECHNOLOGICALLY DEFINED DIGITAL PROPERTYMore and more, technology is being used as a means ofregulating our behaviour in relation to informational products. Inhis seminal book, Code and Other Laws of Cyberspace,375Stanford University law professor Larry Lessig highlights how thedigital environment is not a given, but rather a construction ofcode writers. The “nature” we inhabit in the digital world is oneconstructed through technology and technologists. In Lessig’stheory, there are four modalities of regulation: customary norms;the market; law; and architecture. If I want to stop someone fromspeeding, I can employ the four modalities of regulation byencouraging a customary norm that speeding is bad through:advertising; raising the price of petrol; enacting a law to sayspeeding is an offence; and building a restraining architecturesuch as a mechanical limit in the car or speed bumps. It is assimple as speed bumps. Just as architecture in real space canconstrain our action, Lessig explains architecture in the digitalworld (code) can regulate what we do.Therefore, instead of relying solely on law (e.g., copyright law)to protect my informational value (e.g., software), I shouldconsider what technological mechanisms are available to regulateaccess and use of my informational product. The big players havealready begun this process and we will hear more and more aboutthe role encryption will serve in the distribution of digitalentertainment informational products through a framework ofdigital rights management. And while many advocate thattechnological restraints need to be principled and give balancedaccess to the public—in the way copyright legislation does—thelegislatures have enacted laws like the DMCA in the UnitedStates and the Copyright Amendment (Digital Agenda) Act inAustralia that serve to buttress technological constraints bymaking it a crime to deal in or provide devices that circumventtechnological protection measures. These types of laws, combinedwith code, will make technological protection measures crucial in375. Lessig, supra note 42.

the new environment.For example, the latest version of a popular software productcontains technological or coded restraints that make it verydifficult to copy and load the software on to a second machine.This technological constraint is designed to enforce copyright inthe software. Likewise, DVDs are distributed with anti-copyingCSS encryption, which will only allow them to be accessed throughan authorised player. Some fear that coded restraints couldbecome a law unto themselves and slant protection too far infavour of the property holder.The Reimerdes case highlights the notion of “code as law.”376The thrust of Lessig’s argument, in contrast to that of early digitallibertarians, was that governments should regulate cyberspace tocounteract the negative effects of code.377 He argued that thepublic values inserted into everyday life by legislation and thecommon law (for example the fair use doctrine in copyright law),were at risk in a code based world where private corporationsregulated activity according to their private values.378 The worldof code needed to be infused with public values through law. TheDMCA, in essence, does the opposite, in that it is a law that hasbeen enacted to preserve the power of the code by prohibitingpeople from breaking through or cracking code. In Lessig’sschema, there should now be another law that sits on top of theDMCA that inserts values into the system.Imagine a world where an informational product can bedistributed to the consumer in a way that copying of and access tothe informational product can be severely restricted. Imagine aworld in which chapters of a book can be downloaded or streamed,but not printed or copied. In this type of scenario, or “pay-perview”world, encryption will be used as the technologicalmechanism for limiting use of informational products. A largenumber of people are concerned that such a state of affairs wouldlead to the demise of our current fair use rights to copy a chapterof a book or to read a book in a library. It could be argued that weshould jettison fair use rights and reconceptualize the way publicinterest will be served in a world of code. Perhaps things mightwork better without fair use and that holding on to such a376. See id.377. See id.378. See id.

principle in the digital/crypto world is conservative and dated.However, I am yet to be convinced that I am a dinosaur, or thatthe demise of fair use is a good thing.How do we ensure the code world and its digital property willespouse public values of user and access? The answer to thisquestion requires us to determine the source and location ofvalues inherent to digital property and their continued relevanceand enforceability.VI. THE VALUES DEFINING DIGITAL PROPERTY: DIGITALCONSTITUTIONALISMAs we have seen to this point, digital property takes definitionfrom legislation, contract and technology. The Constitution, as afoundational document bringing fairness to the exercise of power,acts as an overlay. However, U.S. constitutional law and itsfundamental values, such as free speech and equality, are onlyoperative where there is evidence of state action: legislative,administrative or judicial.379 Therefore, U.S. constitutional lawwill have limited application in the privatised world of code, whereprivate technology corporations build self-executing digitalarchitecture embedded/coded with corporate, not public values.The doctrine of state action, which perpetuates the public/privatedivide, is a crucial issue to understand. Some suggest its enduringnature in U.S. constitutional law comes from the notion of laissezfaire—the idea that people like to think their private actions areregulated by little more than the market.380 Whatever itsrationale, the notion of state action means that, in many cases,values inherent in public constitutionalism will have limitedimpact in the digital world. This challenges us to conceptualise anotion of digital constitutionalism wherein public values are379. See Flagg Bros., Inc. v. Brooks, 436 U.S. 149 (1978); Shelley v. Kraemer, 334U.S. 1 (1948); Laurence H. Tribe, American Constitutional Law 1711 (2d ed. 1988); seealso Hudgens v. NLRB, 424 U.S. 507 (1975) (holding that picketers did not have a FirstAmendment right to enter the shopping center for the purpose of advertising theirstrike against their employer); Amalgamated Food Employees Union Local 590 v.Logan Valley Plaza, Inc., 391 U.S. 308 (1968) (holding that a shopping center whichserves as a community business block was the functional equivalent of a business blockfor First Amendment purposes); Andrew Clapham, Human Rights in the PrivateSphere (Clarendon Press, Oxford 1994).380. Paul S. Berman, Cyberspace and the State Action Debate: The Cultural Valueof Applying Constitutional Norms to “Private” Regulation, 71 U. Colo. L. Rev. 1263,1288 (2000).

implemented. As this section will highlight, the notion of digitalconstitutionalism, which seeks to implement a principledallocation of power/rights over information, will combine aspectsof public constitutionalism, statutory and common lawadjudication and perhaps new legislation in relation totechnology/code.Where digital property rights are founded on legislative actionpursuant to the IP clause, then constitutional values inherent inthat clause, as well as notions such as free speech, will need to beweighed in the balance. Furthermore, general law principles ofaccess and user rights, as evidenced by fair use doctrine andantitrust law, will provide definition to any legislated digitalproperty right. I have argued above that legal principles such asantitrust and fair use have the potential to implement/guaranteea diversity ethic in the digital environment.For digital property rights constructed or extended throughcontract, the role of constitutionally entrenched rights and valueswill be more remote. If the contract is simply building onlegislative rights, then the constitutional issues will be relevant tothat aspect, but what of the situation where contract purports tocreate digital property rights, e.g., in the case of raw data. Firstwould come the question of whether this conflicts with theConstitution or legislation, such as section 301 of the CopyrightAct; this might require further consideration as to whether it is anegotiated or mass market licence, the latter being more likeconflicting legislative action.381 However, assuming we take theview enunciated in ProCD that such a contractual process islawful, then we are left to find values in the field of contract law.A vigorous notion of unconscionability developed to address thespecific needs of information contracts might provide a way ofensuring that access and user rights are maintained. However, ifcontract is tied with technology and the contract is self-executingthrough technology—meaning access to data is denied if copyingfor fair use occurs—then enforcement of such a principle willrequire the user to litigate and enforce an emerging principle of381. Custom or negotiated software licences are much more specific andcontractual in nature. Mass market or non-negotiated licences, on the other hand,often act like a multilateral treaty or legislation to guarantee a uniform customamongst all users. Some argue the latter are being preempted due to their legislativelikecharacter while the former are more contractual in nature and therefore notpreempted.

unknown quality i.e., informational unconscionability.382 Thepotential of self-executing contracts would also mean that stateaction would be hard to establish against the property holder andtherefore issues such as free speech would not easily be raised.A further question concerns the extent to which contract canbe used to oust fair use rights bestowed by the Copyright Act.383Does contract have the power to define digital property byremoving access and user rights granted by the Copyright Act toengage in reverse engineering of software for interoperabilitypurposes? This issue is hotly contested by different interestgroups; however, if contract does have this capacity, then itbecomes a most important strategy in defining digital propertyrights. Another issue will be the extent to which contract islimited by antitrust doctrine through its definition of digitalproperty.Furthermore, can contract oust the first sale doctrine? Thefirst sale doctrine, expressed in section 109 of the U.S. CopyrightAct384 and implied in Australian copyright law,385 allows thecopyright owner to control distribution of the tangible embodimentof copyright up to the point of first sale. For example, if I buy abook, I gain an ownership right to the book and can sell it but notreproduce it, unless I have a fair use/fair dealing right. I gainownership of the tangible book and thereby reduce the power ofthe copyright owner, but I do not gain ownership of the remainingexclusive rights of the copyright owners such as the reproduction382. See J.H. Reichman & Jonathon A. Franklin, Privately Legislated IntellectualProperty Rights: Reconciling Freedom of Contract with Public Good Uses ofInformation, 147 U. Pa. L. Rev. 875 (1999).383. See UCITA § 105 (1999); Brian F. Fitzgerald, Commodifying and TransactingInformational Products Through Contractual Licences: The Challenge for InformationalConstitutionalism, in Intellectual Property and the Common Law World 35 (CharlesE.F. Rickett & Graeme W. Austin eds., 2000).384. See also 17 U.S.C. § 202 (1994).Ownership of a copyright, or of any of the exclusive rights under a copyright,is distinct from ownership of any material object in which the work isembodied. Transfer of ownership of any material object, including the copy orphonorecord in which the work is first fixed, does not of itself convey anyrights in the copyrighted work embodied in the object; nor, in the absence ofan agreement, does transfer of ownership of a copyright or of any exclusiverights under a copyright convey property rights in any material object.Id.385. See Pacific Film Laboratories Pty. Ltd. v. Fed. Comm’r of Taxation (1970) 121C.L.R. 154 (Austl.).

ight. First sale is an important distributive principle and to someextent software licensing challenges this distributive mechanism.Software is generally licensed, not sold, as is an ever increasingvariety of information products. We are sold access or user rightsto information rather than tangible ownership rights. This isdone through a contractual user agreement/licence, whichstipulates the user rights of consumers. The contractual licenceallows greater control over the exploitation of the informationalproduct. On one hand, it could be that contract cannot be used tooverride such a fundamental principle of copyright law.Informational products are either sold or leased. In DSCCommunications Corp. v. Pulse Communications, Inc.,386 the courtsuggested that contract can be used to oust first sale.387Lastly, there is technologically created property. Once again,code has the power to define and enforce, through self-execution,access and user rights in accordance with the values of the privatecorporation setting the code. In this situation, the Constitution isunlikely to provide the values as there is no state actor, and inmany cases, no need to use a court for enforcement purposes.However, if contract is intertwined with the coded distribution ofinformation, there may be points at which constitutional orgeneral law values can be argued. In the absence of applicableconstitutional doctrine, statutory law or common law, technologywill be used to construct property rights in the vision of theprivate corporation. This is the point at which Lessig suggests weneed to legislate for digital constitutional values. Likewise,Berman has suggested that we need to invoke principles of publicconstitutionalism in disputes emanating from the privatesphere.388It is worth citing two examples to give further insight onthese issues. There is the case where a large ISP blocks outunsolicited e-mail sent by a marketing firm (spam), who thenargues that such action is an infringement on free speech.389 The386. 170 F.3d 1354 (Fed. Cir. 1999).387. See id. at 1361.388. See Berman, supra note 380, at 1290 (calling this notion ‘constitutiveconstitutionalism’: the idea that the Constitution might appropriately be viewed as thetouchstone for articulating constitutive values and for structuring public debate aboutfundamental social and political issues).389. Cyber Promotions, Inc. v. Am. Online, Inc., 948 F. Supp. 436, 438 (E.D. Pa.1996).

ISP explains that they are not a state actor and the courtagrees.390 Should private action of such a public nature, wherethe ISP’s mailboxes and networks are vital to the widespreaddissemination of information, trump the marketing firm’s freespeech interest? Should the private property be open for thispurpose? In general, owners of private property are not requiredto allow their property to be used for speech purposes thoughthere are exceptions. Is there a principle of free speech in thisprivate space where property law and notions such as trespass391and nuisance predominate? Had the ISP sought to enforce itsrights through a trespass action, the issue of free speech and stateaction may gain more support.392 However, through selfexecutingcode, an ISP is able to enforce its private norms, forcingthe marketing firm to bring the action to court. In doing so,marketing firms must scramble to find an unconscionability ordiversity principle/ethic in their favour since such a principle hasyet to be fully developed for the digital environment. The lack of asophisticated notion of digital constitutionalism hinders themaking of an already difficult argument.Secondly, imagine a private corporation that encrypts anddistributes information it on its own terms. UCITA contains thepower to be activated remotely or through self-help;393 thus it isself-enforcing with no need for the court system. Code rules. Inthis case, the values or principles of the transaction are insertedand implemented by the private corporation. The public values offair use and free access, and in a broader sense respect for thepublic domain, are not necessarily implemented in this regime.For example, I buy a book that may not be printed or copied andmay be read or viewed only once. The code enforces this processand therefore a court will not be brought into the question unlessthere is a valid cause of action, i.e., unconsionability.Contrast contract outside the realm of code. While I might390. Id. at 445391. See e.g., eBay, Inc. v. Bidder’s Edge, Inc., 100 F. Supp. 2d. 1058 (N.D. Cal.2000) (involving a case where an Internet auction site sued an auction aggregating sitefor trespass).392. Cf. Compuserve, Inc. v. Cyber Promotions, Inc., 962 F. Supp. 1015 (S.D. Ohio1997) (holding that a company’s intentional use of the service’s proprietary computerequipment was actionable trespass to chattels for which First Amendment provided nodefense).393. See UCITA § 816 (1999); Julie E. Cohen, Copyright and the Jurisprudence ofSelf-Help, 13 Berkeley Tech. L.J. 1089 (1998).

sign a hard copy contract containing what I consider to be thecomplete terms of the agreement, the legislature and the courtsmay still insert public interest principles, such as the notion ofunconscionability. Do we need legislation to ensure that thesenew code driven information contracts are informationallyconscionable? Will the current law suffice?Lessig explains much the same thing in the following way:The dissimilarity is this: with every enforced contract—with every agreement that subsequently calls upon anenforcer to carry out the terms of that agreement—thereis a judgment made by the enforcer about whether thisobligation should be enforced. In the main, thesejudgments are made by a court. And when a court makessuch judgments, the court considers not just the privateorderings constituted in the agreement before it, but alsoissues of public policy, which can, in some contexts,override these private orderings. When a court enforcesthe agreement, it decides how far the power of the courtcan be used to carry out the agreement. Sometimes theagreement will be carried out in full; but often, theagreements cannot be fully effected. Doctrines such asimpossibility or mistake will discharge certainobligations. Rules about remedy will limit the remediesthe parties can seek. Public policy exceptions willcondition the kinds of agreements that can be enforced.‘Contracts’ incorporate all these doctrines, and it is themix of this set of public values, and private obligations,that together produce what we call ‘a contract.’When the code enforces agreements, however, or whenthe code carries out a self-imposed constraint, thesepublic values do not necessarily enter into the mix.Consequences that a court might resist (forfeitures, forexample), the code can impose without hesitation. Thecode writer operates free of the implicit limitations ofcontract law. He or she can construct an alternativeregime for enforcing voluntary constraints. And nothingrequires or ensures that this alternative regime willcomport with the values of the background regime we call‘contract.’

. . . .For again, in real space, one might well believe that aset of obligations imposed through contract wasuntroubling. Conditioned by antitrust law, limited byprinciples of equity, cabined by doctrines of mistake andexcuse—the obligations would be checked by a courtbefore the constraints were made effective. There is astructural safety check on obligations of this sort, whichensures that the obligations don’t reach too deep. Whenintervening to enforce these obligations, a court wouldcarry with it the collection of tools that contract law hasdeveloped to modify, or soften, the obligations thatcontract law might otherwise enforce.The cyberspace analog has no equivalent toolbox. Itsobligations are not conditioned by the public values thatcontract law embraces. Its obligations instead flowautomatically from the structures imposed in the code.These structures serve the private ends of the codewriter; they are a private version of contract law. But asthe Legal Realists spent a generation teaching, and as weseem so keen to forget: contract law is public law.394The point to make is that technology will play a significantrole in defining digital property and it is imperative that thevalues of such technologically constructed digital property beclearly enunciated. These are values that will underpin theconstruction of culture and knowledge, and define the frameworkfor the creation and transfer of wealth through digitisedinformational products.CONCLUSION: THE ULTIMATE BOUNDARY? A QUESTION OF DIGITALCONSTITUTIONALISMThe purpose of this article has been to seek a betterunderstanding and definition of digital property. As the foregoingpages show, this is a complex process of reasoning.At a more traditional constitutional level, the intellectualproperty enumerating clauses will set limits and bring definition394. Lawrence Lessig, The Law of the Horse: What Cyberlaw Might Teach, 113Harv. L. Rev. 501, 530-31 (1999) (footnotes omitted).

to digital property. As well, free speech, where constitutionallyentrenched, has the potential to rigorously refine digital property,but has yet to fully achieve this outcome.One of the most crucial points to appreciate in defining digitalproperty is the role of concepts of statutory or common law (nonconstitutionallaw). As the discussion of antitrust and fair useprinciples highlights, general law will play an aggressive role indefining the boundaries of digital property and has the potentialto infuse it with a diversity ethic designed to facilitate distributedintelligence and democratic society.Next, we need to understand how contract and technology willcoalesce in the digital world to refine access and user rights—toimplement the emerging pay-per-view world/culture in which fairuse rights evaporate in the face of information user rightsdistributed through digital rights management systems rooted inencryption and reinforced by anti-circumventions laws. Facedwith this emerging technology based distribution of digitalproperty, we are forced to ask whether “any” information at allshould be open to be freely accessed or copied? If we answer “yes,”then we must advocate digital rights that enhance these socialattributes. This challenges us to propose principles of digitalconstitutionalism: on one hand, we must question howinformational power should be exercised and on the other, howvalue should be protected? For many of us, complete ownershipwill not be appreciated. A negotiated multitudinous distributednotion of ownership is sought in which the cultural dynamic ofinformation must be appreciated.In the final reckoning, the ultimate boundary of digitalproperty will remain a hotly contested legal concept in whichmany aspects of law and culture will be presented. The more werealise the inherent and incredible discursive nature ofinformation and informational products, the more we will look fordiversity in the construction and use of digital property, as well asthe principles that will guarantee this diversity against abackdrop of what might be termed “digital constitutionalism.”

What the Future Holds: PolicyChoices in the Global E-MarketplaceMaureen A. O’Rourke *This essay’s title is a misnomer. No one knows “What theFuture Holds” in the global marketplace. We can, however, makesome educated guesses about the issues that will assumeparticular importance in the next year or so. Thus, the essaymight be more accurately titled, “What May or May Not Be theImportant Issues in Internet Commerce in the Next Year or So”—less catchy, admittedly, but closer to the truth.I suspect that policymakers will spend much of their timefocusing on three related questions: (i) access (in at least threedifferent forms); (ii) how to encourage e-commerce generally and(iii) whether e-commerce in information products implicatesconcerns different from those arising in the sale of goods online.This essay reviews issues under each one of these categories. Thegoal is primarily to identify the debates rather than to suggestsolutions. E-commerce raises complex issues not amenable tosimple solutions in an essay and, indeed, that will require carefulstudy over time.I. QUESTIONS OF ACCESSE-commerce raises questions of access in a number ofdifferent forms. First, e-tailers generally would like to offer theirproducts to the largest number of potential purchasers possible. Ifmany consumers do not have access to the technology that allowsthem to connect to the Internet in an effective way, the marketwill be smaller than it otherwise would be, and therefore, less*Professor of Law, Associate Dean for Administration & Graduate Programs, MichaelsFaculty Research Scholar, Boston University School of Law. Many thanks to DavidRice for inviting me to the symposium, Information and Electronic Commerce Law:Comparative Perspectives, and to my research assistant, Casey Caldwell, for his help.237

desirable to sellers and buyers alike. Moreover, consumerswithout access to the Internet will miss not merely onlineshopping opportunities, but also access to the mass of sometimesedifying information available on the Web.This problem has at least two related facets: who will haveaccess to technology and what specific means of access may theylegitimately employ? Policymakers have labeled the formerproblem of unequal access to technology the “Digital Divide.”1Indeed, this issue has been with us for so many years that theterm “Digital Divide” has grown in the collective consciousness toa point where it is almost passé. Nevertheless, unequal access toInternet technology will remain a concern in the near future andinequalities may grow for some time rather than shrink. Even ifpolicymakers could solve the problem of unequal opportunity toaccess the Internet, what technology would the formerlydisenfranchised or even the Internet-savvy use? As technologymarches on, the question of who will have access to the bettertechnology, like broadband, arises.While almost everyone is familiar with at least some of theissues the “Digital Divide” raises, fewer may know about an accessdebate currently raging in the United States. That debate focuseson who may link to a Website and how they may do so. In recentmonths, certain popular sites have begun to employ an ancienttort, trespass to chattels, in an attempt to regulate access to theirsites. Depending on how courts finally resolve such claims, theuse of trespass to chattels to regulate access to sites could hamperthe emergence of the Internet as a market that approaches perfectcompetition and promises significant benefits for consumers.The following discussion addresses problems of access. Itdiscusses some parts of the debate over the “Digital Divide” andaccess to broadband technologies. It then reviews some of thecaselaw on access to Websites and proposes an approach toresolving such questions.A. Problems of Access Generally—The Digital DivideCommentators have thoroughly documented the “Digital1. Report of the Digital Opportunity Task Force, Digital Opportunities for All:Meeting the Challenge, at 3 (May 11, 2001), available athttp://www.glocom.ac.jp/dotforce/final/DOTForceReportv50g.pdf.

Divide” both domestically and outside of the United States.2Within the United States, there are dramatic disparities in accessto technology across income levels.3 Geography makes adifference as well: rural Americans are less likely to have access totechnology than their urban counterparts.4 Indeed, less than fivepercent of towns with populations under 10,000 have access tobroadband technology.5 Poor, rural Americans are twenty timesmore likely to be left behind in matters of technology than theirfellow citizens.6 Minorities and the disabled are less likely to owncomputers.7 These disparities persist despite efforts by the publicand private sectors to narrow the gap.8In May 2001, the Digital Opportunity Task Force (DOT Force)released its proposals for bridging the global gap. The G8 heads ofstate had created the DOT Force at the Kyushu-Okinawa summitin July 2000 “in a cooperative effort to identify ways in which thedigital revolution can benefit all the world’s people, especially thepoorest and most marginalized groups.”9 According to the DOTForce Report:One third of the world population has never made atelephone call. Seventy percent of the world’s poor live in2. See U.S. Dept. of Com., Leadership for the New Millennium—Delivering onDigital Progress and Prosperity, 3d Annual Report, 3-19 (2000) [hereinafter Deliveringon Digital Progress] (discussing the Clinton Administration’s initiatives in providingdirect government assistance to the technologically disadvantaged, and in encouragingthe private sector to do the same).3. See Henry H. Perritt, Jr., Economic and Other Barriers to ElectronicCommerce, 21 U. Pa. J. Int’l Econ. L. 563 (2000); Bob Rowe, Strategies to PromoteAdvanced Telecommunications Capabilities, 52 Fed. Comm. L.J. 381 (2000).4. See Rowe, supra note 3, at 386; see also J.M. Spectar, Bridging the GlobalDigital Divide: Frameworks for Access and the World Wide Wireless Web, 26 N.C.J. Int’lL. & Com. Reg. 57 (2000) (discussing the need to institute policies designed to easebarriers to Internet usage among persons living in rural communities).5. Molly M. Peterson, Broadband Battle Over Small-Town U.S.A., Nat’l J., June16, 2001, available at 2001 WL 7182298.6. Gephardt Statement on the Democratic E-Strategy for Economic Growth (2/6),U.S. Newswire, Apr. 5, 2001, available at 2001 WL 4141881, at *3.7. See Margaret Chon, Erasing Race? A Critical Race Feminist View of InternetIdentity-Shifting, 3 J. Gender Race & Just. 439 (2000); Jerry Kang, Cyber-Race, 113Harv. L. Rev. 1130 (2000).8. See generally William E. Kennard, Equality in the Information Age, 51 Fed.Comm. L.J. 553 (1999) (discussing the disparity in obtaining Internet access based onrace and income, and the public and private sectors’ efforts to bridge the gap); Mark J.Maier, Affordable Internet Access for All Americans, 6 Rich. J.L. & Tech. 8 (1999)(discussing how government incentives can be used to encourage affordable pricing).9. Digital Opportunity Task Force, supra note 1, at 3.

ural and remote areas, where access to information andcommunications technologies [ICT], even to a telephone,is often scarce. Most of the information exchanged overglobal networks such as the Internet is in English, thelanguage of less than ten percent of the world’spopulation.10Those statistics, while grim, are not surprising. As the Reportrecognizes, they reflect the general socio-economic disparitybetween the developed and less-developed world.11The DOT Force Report sets forth nine action points as itsGenoa Plan of Action.12 They are:1. Help Establish and Support Developing Country &Emerging National e-Strategies;2. Improve Connectivity, Increase Access and Lower Costs;3. Enhance Human Capacity Development, KnowledgeCreation and Sharing;4. Foster Enterprise and Entrepreneurship for SustainableEconomic Development;5. Establish and Support Universal Participation inAddressing New International Policy and TechnicalIssues raised by the Internet and ICT;6. Establish and Support Dedicated Initiatives for the ICTInclusion of the Least Developed Countries;7. Promote ICT for Health Care and in Support AgainstHIV/AIDS and Other Infectious and CommunicableDiseases;8. National and International Effort to Support LocalContent and Applications Creation;9. Prioritize ICT in G8 and Other Development AssistancePolicies and Programmes and Enhance Coordination of10. Id. at 6.11. See id.[The] “digital divide” is, in effect, a reflection of existing broader socioeconomicinequalities and can be characterized by insufficient infrastructure,high cost of access, inappropriate or weak policy regimes, inefficiencies in theprovision of telecommunication networks and services, lack of locally createdcontent, and uneven ability to derive economic and social benefits frominformation-intensive activities.Id.12. Genoa was the site of the G8 meetings from July 19-22, 2001.

Multilateral Initiatives.13The DOT Force places responsibility for implementing its planwith developing countries’ governments working alongsidecorporations and non-governmental organizations.14 It plans toplay an advisory role in assisting these entities in taking therelevant steps toward bridging the “Digital Divide.”15Reducing a problem as complex as the global “Digital Divide”to a nine-step plan of action is indeed a noteworthyaccomplishment. However, skeptics might question whether—even if the DOT Force plan would work—the money and willexists to implement it. Translating the plan from politicalplatitudes into action requires money, and the Report says littleabout from what sources and in what magnitude that money willcome. Further, the Report makes certain assumptions about thedesirability of competitive markets that some nations do notshare.16 Developing countries may not be interested in the GenoaPlan if they view it as simply another way in which the developedworld seeks to foist its policy agenda on less-developed countries.Indeed, the most difficult part of implementing any global planmay be determining how to accommodate local cultural valuesthat can vary widely, even within a single country.B. Access to Broadband TechnologiesDecisions of policymakers will also influence what technologycitizens have access to, and under what terms. Technologymarches on, but only about five percent of American householdshave access to broadband.17 One definition describes broadband13. See Digital Opportunity Task Force, supra note 1, at 13-20.14. See id. at 10.15. See id. (“[T]he DOT Force can . . . play a critical and significant role bysuggesting, initiating and/or supporting these actions.”).16. See id. at 8 (stating that “pro-competitive policies” are prerequisites to thedevelopment of communications infrastructure and access, which are required toreduce the digital divide).17. The Internet Freedom and Broadband Deployment Act of Law: Hearing onH.R. 1542 Before the House Comm. on the Judiciary, 107th Cong. (2001), available at2001 WL 21755660, at *3 (statement by James K. Glassman, Resident Fellow,American Enterprises Inst., House Comm. on the Judiciary) (“Today, barely 5 percentof households have even the most rudimentary form of broadband—or fast access to theInternet. As a result, the promise of the Internet is not being enjoyed by 95 percent ofAmericans—an increasingly frustrated group. Incredibly, in the first quarter of thisyear, for the first time ever, the number of Americans with Internet access of any sort

as “high-speed telecommunications capability” that “requires botha high capacity backbone with adequate access points, and localdistribution from those points of access to the users.”18 Generallyproviders have offered three technologies to provide this service:(i) DSL (Digital or Direct Subscriber Line); (ii) cable modem accessand (iii) wireless access. Although each technology can providebroadband access, the law regulates them under different regimes.For example, a common carrier/open access regulatory modelgoverns DSL. This technology uses the telephone system; andprinciples of open access bind the Regional Bell OperatingCompanies (RBOCs).19 In contrast, no such open access modelapplies to cable companies. Instead, such companies are largelyunregulated. Some providers of cable modem service offerInternet access only through an affiliated Internet ServiceProvider (ISP). This has led to calls for principles of open accessto govern cable firms as well.20 However, the question of how toregulate cable companies (if at all) is complex.It is unclear under what statutory category cable modemservices fall, or which agency has the regulatory authority togovern such services. Indeed, two courts have reached oppositeconclusions about whether or not Internet transmission acrosscable lines constitutes a “cable service.”21 In 2000, the FederalCommunications Commission (FCC) issued a Notice of Inquiryrequesting comments on whether it should treat cable modemservice as a cable service, a telecommunications service, aninformation service, or some hybrid subject to a number of theactually declined. As Francis Rose wrote recently in Wired magazine . . . ‘The digitalfuture has arrived, but the analog past won’t let go. Data flashes across the country atthe speed of light only to end up dribbling out of your wall in the tech version ofChinese water torture.’”).18. Delivering on Digital Progress, supra note 2, at 59.19. See Julian Epstein, A Lite Touch on Broadband: Achieving the OptimalRegulatory Efficiency in the Internet Broadband Market, 38 Harv. J. on Legis. 37 (2001)(discussing the RBOCs’ argument for open access to cable systems for ISPs, in thecontext of the RBOCs’ obligation to allow open access to their own networks under theTelecommunications Act of 1996).20. See Douglas G. Lichtman et. al., Telecommunications Law and Policy (2001).21. Compare AT&T v. City of Portland, 216 F.3d 871 (9th Cir. 2000) (holding thatInternet transmission over cable is a telecommunication service, not a cable service,within the meaning of the Communications Act and that that Act prohibits localgovernments from imposing open access requirements) with MediaOne v. County ofHenrico, 97 F. Supp. 2d 712 (E.D. Va. 2000) (labeling such Internet transmission a“cable service” and holding that local regulation is preempted).

Communication Act’s provisions.22 Regardless of what statutorybasket governs cable modem service, the case for open access isitself uncertain.The FCC initially chose to pursue a hands-off policy towardscable modem service.23 Then FCC Chairman William Kennarddescribed the FCC’s stance as “first, do no harm. Call it a hightechHippocratic Oath.”24 This policy stance reflected a fear thatimposing regulations before the broadband market developedwould deter that market’s growth, in part because the investmentrequired to establish such systems is quite large.25 Thus, the FCCimposed no open access requirement in the mergers of AT&T andTCI or AT&T and MediaOne.26 In the latter case, the FCCbelieved that it should not disturb the deployment of alternativetechnology as it was occurring.27 Moreover, the company hadcommitted itself to negotiating non-exclusive licenses withunaffiliated ISPs when its exclusive arrangements with affiliatedISPs expired.28Open access itself is difficult to define. In its Notice ofInquiry, the FCC noted that “[m]ost open access proposals entailtwo broad requirements, providing unaffiliated ISPs with theright to (i) purchase transmission capacity; and (ii) access thecustomer directly from the incumbent cable operator.”29 The mostobvious potential benefit of open access is that it may help toprevent both discrimination against an unaffiliated ISP’s contentby the dominant service provider and a monopoly that might22. See Inquiry Concerning High-Speed Access to the Internet Over Cable andOther Facilities, Notice of Inquiry, GN Docket No. 00-185, FCC 00-355 (Sept. 28, 2000),available at http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-00-355A1.pdf; seealso Chairman William E. Kennard, Statement Concerning Notice of Inquiry IntoHigh-Speed Internet Service (Sept. 28, 2000), available athttp://ftp.fcc.gov/Speeches/Kennard/Statements/2000/stwek077.html.23. See Notice of Inquiry, supra note 22, at para. 4.24. Lichtman, supra note 20, at 26 (quoting Chairman William E. Kennard, TheRoad Not Taken: Building a Broadband Future for America, Remarks Before theNational Cable Television Association, Chicago, Ill. (June 15, 1999)).25. See id.26. See id. at 35 (citing paragraph 12 of the Notice of Inquiry). In contrast, theFTC conditioned approval of the AOL-Time Warner merger on the combined company’sagreement to provide open access to ISPs. See Daniel L. Rubinfield & Hal J. Singer,Open Access to Broadband Networks: A Case Study of the AOL/Time Warner Merger,16 Berkeley Tech. L.J. 631 (2001).27. Lichtman, supra note 20, at 35.28. See Notice of Inquiry, supra note 22, at para. 12.29. Id. at 38 (quoting paragraph 27 of the Notice).

decrease innovation.30 On the other hand, despite some exclusivearrangements, it is not clear that cable modem service providerswill gravitate toward a model that restricts consumer choice ofISPs. In other words, the market may function quite well. If thereis value in offering consumers a choice of ISPs, cable companieswill do so in order to command a higher price for their Internetconnections. Moreover, open access may actually decreaseincentives to innovate by decreasing the returns that accrue to thecable company. By involving a larger number of parties indeveloping the infrastructure, open access may enhance thechances that the industry will standardize on the best technology.On the other hand, open access may slow the development processbecause the involvement of many decisionmakers can lead todelays.Finally, wireless technology raises the question of howpolicymakers should allocate spectrum.31 The FCC assigned partsof the spectrum in the 1960s, giving large chunks to government(especially the Department of Defense) and educationalinstitutions.32 Reallocating those assignments will lead topolitical battles among competing interest groups.The World Radiocommunication Conference of 2000 endorsedefforts to “harmonize spectrum allocations regionally andinternationally.”33 A staff-level report from the FCC andDepartment of Commerce concluded, however, that seriousobstacles would hamper the United States from clearing bandsthat the Conference highlighted as suitable for third generation(3G) uses.34 Educational institutions and others already usemany of these bands, raising political, technical and economicquestions about the feasibility of re-allocating or sharing that30. Id. at 48.31. See Delivering on Digital Progress, supra note 2, at 62-63 (noting the U.S.government’s endorsement of the principles adopted by the World RadiocommunicationConference of 2000: “(1) governments may choose spectrum from any one or all of thebands identified for third generation mobile wireless (3G); (2) governments have theflexibility to identify spectrum if and when they choose; and (3) no specific technologywill be identified for third generation services,” and noting its call for support ofindustry efforts to harmonize spectrum allocations regionally and internationally); seealso Anthony Shadid, Fight Puts Airwaves at Risk, Boston Globe, Feb. 10, 2001, at A1.32. Shadid, supra note 31, at A1.33. Delivering on Digital Progress, supra note 2, at 63.34. See Aaron Pressman, White House Undercuts Spectrum Reports, The IndustryStandard, Mar. 30, 2001, available atwww.thestandard.com/article/0,1902,23258,00.html.

spectrum.35 However, the Bush administration is reportedlyseeking to distance itself from these conclusions, thus raisingquestions about what its wireless strategy will be.36The debate over how to regulate broadband technology willlikely continue as Congress considers the “Internet Freedom andBroadband Deployment Act of 2001.”37 This Act, developed inresponse to congressional concerns about impediments toconsumer access to broadband technology, aims to deregulate thebroadband services industry and promote the availability ofchoices for private users of the technology.38C. Access to Publicly Available WebsitesTwo recent cases raise the question whether a Website canprevent another from linking to it by automated means, copyingits information and extracting uncopyrighted product and pricinginformation.39 In Ticketmaster Corp. v. Tickets.com, Inc.,40 theFederal District Court for the Central District of Californiainitially dismissed Ticketmaster’s state law claim that unwantedlinking constituted a trespass to its Website, holding that theclaim was preempted by the federal Copyright Act.41 The courtnoted: “it is hard to see how entering a publicly available website35. See id.36. See id. (reporting Commerce Secretary Don Evans’s meeting with majorwireless carriers at which he dismissed the conclusions of the report).37. H.R. 1542, 107th Cong. § 1 (2001).38. See id. at § 2.39. See also Register.com, Inc. v. Verio, Inc., 126 F. Supp. 2d 238 (S.D.N.Y. 2000)(adopting eBay’s trespass theory and also upholding a contract where assent to itsterms was indicated by submitting a query).40. 54 U.S.P.Q. 2d 1344, 1345 (C.D. Cal. 2000) (No. 99-7654-HLH (BQRx)).41. Ticketmaster asserted a variety of causes of action against Tickets.com forsearching the Ticketmaster site, taking its information, providing that information toTickets.com’s visitors, and allowing such visitors to deep-link to the Ticketmaster site.See First Amended Complaint for Injunctive Relief and Damages, Ticketmaster Corp.v. Tickets.com, Inc., No. 99-7654 HLH (BQRx) (C.D. Cal. 2000). The district courtrefused to dismiss the complaint. See Ticketmaster Corp. v. Tickets.com Inc., 54U.S.P.Q. 2d 1344, 1345 (C.D. Cal. Mar. 27, 2000). Thereafter, Ticketmaster filed anamended complaint. See Second Amended Complaint for Injunctive Relief andDamages, Ticketmaster Corp. v. Tickets.com, Inc., No. 99-7654 HLH (BQRx) (C.D. Cal.2000). The district court refused to grant a preliminary injunction against Tickets.combased on this amended complaint. See Ticketmaster Corp. v. Tickets.com, Inc., 2000U.S. Dist. LEXIS 12987. The Ninth Circuit upheld this denial without opinion. SeeTicketmaster Corp. v. Tickets.com, Inc., 2001 WL 51509 (9th Cir. 2001).

could be called a trespass, since all are invited to enter.”42Ticketmaster amended its complaint to reword its claim astrespass to the computer system rather than to the Website. Thecourt rejected this claim as well, noting that Ticketmaster had notshown the functioning of its computer systems to be obstructed.43In contrast, in eBay, Inc. v. Bidder’s Edge, Inc.,44 anotherfederal district court in California held that a linker’s conductcould constitute a trespass to chattels because of the prospect offuture harm that could occur if many linkers chose to visit the siteat the same time.45 The eBay court also held that any access(including that of an individual linking to a site) necessarilyimposes a burden on the visited site’s system, justifyingapplication of trespass law.46Interestingly, the eBay court effectively formulated a newtort, despite grounding its holding on trespass to chattels. Earliercourts had held that electronic signals are tangible enough toconstitute an invasion sufficient to sustain an action fortrespass.47 However, no court had applied a strict liabilitystandard to such an invasion. Trespass to real property lawrequires that the plaintiff show actual harm when the intrusion isintangible.48 Alternatively, some courts have held that theappropriate cause of action for an intangible intrusion is nuisancerather than trespass.49 Nuisance law weighs the costs and42. Ticketmaster, 54 U.S.P.Q. 2d at 1345.43. See Ticketmaster Corp. v. Tickets.com, Inc., No. CV99-7654, 2000 U.S. Dist.LEXIS 12987, at *17 (C.D. Cal. Aug. 10, 2000) (unpublished minute order).44. 100 F. Supp. 2d 1058 (N.D. Cal. 2000).45. Id. at 1069; see also First Amended Complaint, eBay, Inc. v. Bidder’s Edge,Inc., No. C-99 21200 (N.D. Cal. 2000) (alleging that Bidder’s Edge’s conduct in using anautomated tool to search the eBay site, copy its information, and extract product andpricing data should afford eBay a remedy under a variety of causes of action). Thedistrict court entered a preliminary injunction against Bidder’s Edge under a trespassto chattels theory. See eBay, 100 F. Supp. 2d at 1069. The parties later settled thesuit. See EBay, Bidder’s Edge Settle Suits on Web Access, L.A. Times, Mar. 2, 2001, atC2. Bidder’s Edge has ceased operation. See A Message to Our Users, athttp://www.biddersedge.com (explaining that Bidder’s Edge would cease operation onFeb. 21, 2001, because of “market and financial conditions”) (last visited Feb. 19, 2001).46. See eBay, 100 F. Supp. 2d at 1071-72.47. See, e.g., Thrifty-Tel, Inc. v. Bezenek, 54 Cal. Rptr. 2d 468 (Cal. Ct. App. 1996)(holding electronic signals used to gain unauthorized access to a computer to betangible enough to support a trespass claim).48. See, e.g., Zaslow v. Kronert, 176 P.2d 1, 7 (Cal. 1946).49. See, e.g., San Diego Gas & Electric Co. v. Superior Court, 920 P.2d 669 (Cal.1996).

enefits of the particular activity when deciding whether or not tohold the invader liable.50 Trespass to chattels, in contrast totrespass to real property, has always required the plaintiff toplead and prove actual harm, not possible future harm.51 TheeBay court, by holding the defendant liable for a harmlessintangible intrusion, thus invented a new cause of action.Developing such a new tort warranted a much more detailedpolicy analysis than that in which the court engaged. Forexample, the court’s economic analysis did not considercompetitive concerns such as the benefit to consumers of easilyavailable product and pricing information. Nor did it discuss howits rule fits with copyright law’s refusal to protect suchinformation or the First Amendment’s protection of commercialspeech. The court also did not consider the nature of the Web. Noone posts an Internet site without expecting—indeed, inviting—some measure of linking from other sources. Whether a siteshould be able to control who links to it and how they do so is nota question amenable to a simple answer like the strict liabilityregime the eBay court adopted.Elsewhere, I have suggested different approaches thatpolicymakers could take to address unwanted linking.52 Courtscould follow traditional trespass and nuisance law, adopting anuisance balancing test to address the intangible invasion thatoccurs when a link is employed. Because balancing tests alwayscreate uncertainty, a statutory “safe harbor,” defining permittedand forbidden means of linking and the acceptable burden a linkermay place on a server, could be created. A linker, though, shouldstill have an opportunity to challenge the safe harbor when theplaintiff’s site has engaged in misconduct. A database bill similarto one already proposed could be easily modified to adopt such anapproach, providing the linker with defenses like misuse to help50. See Restatement (Second) of Torts § 831 (1965).51. See Restatement (Second) of Torts § 218 (1965).52. See Maureen A. O’Rourke, Property Rights and Competition on the Internet: InSearch of an Appropriate Analogy, 16 Berkeley Tech. L.J. 561 (2001) (discussing howcompetition policy and antitrust law can aid policymakers in deciding how to defineproperty rights on the Internet) [hereinafter O’Rourke, Property Rights]; Maureen A.O’Rourke, Shaping Competition on the Internet: Who Owns Product and PricingInformation?, 53 Vand. L. Rev. 1965 (2000) (reviewing the competitive environment onthe Internet and the causes of action in the eBay case) [hereinafter O’Rourke, ShapingCompetition].

inform what permissible access and means of access may be.53II. ENCOURAGING E-COMMERCEEven if all consumers could have effective access to Internettechnology and efficient access to Websites, e-commerce may notreach its potential. Both consumers and online retailers haveconcerns that may lead them to limit their online activities. Thefollowing briefly summarizes some of those issues.If consumers cannot trust e-tailers, they will not shop online.Indeed, Forrester Research estimates that consumer spendingonline would have been $12.4 billion higher in 1999 if consumershad not been worried about the privacy of their information.54Those concerns are exacerbated by uncertainty over the treatmentof customer information in the event of bankruptcy proceedings.In May 2000, Toysmart, an e-tailer of children’s toys, wentbankrupt and sought to sell customer information that it collectedunder a privacy policy promising that such information wouldnever be sold.55 Although the case was resolved with thepurchase and destruction of the list by Toysmart’s major investor,its practical impact may be that privacy policies will now explicitlyprovide for the sale of customer information. Amazon and eBayadjusted their privacy policies in the aftermath of the Toysmartcase. Those policies now provide for the transfer of customerinformation to third parties in certain circumstances.5653. See O’Rourke, Property Rights, supra note 52; O’Rourke Shaping Competition,supra note 52; see also Dan L. Burk, The Trouble With Trespass, 4 J. Small &Emerging Bus. L. 27 (2000) (arguing for the adoption of a nuisance standard to addressunwanted access to websites).54. See Anthony Shadid, Crackdown Seen on Customer Databases, Boston Globe,Jan. 8, 2001, at C1.55. See Andrew B. Buxbaum & Louis A. Curcio, Note, When You Can’t Sell toYour Customers, Try Selling Your Customers (But Not Under the Bankruptcy Code), 8Am. Bankr. Inst. L. Rev. 395 (2000) (arguing that the Bankruptcy Code should beinterpreted to prohibit the sale of customer information gathered by a bankruptWebsite when the site promises not to transfer the information to a third party).56. See Tamara Loomis, Amazon Revamps its Policy on Sharing Data, N.Y.L.J.,Sept. 21, 2000, at 5 (describing Amazon’s change of policy from one promising not toshare information except potentially to other “trustworthy third parties” to one thatlists circumstances in which Amazon shares information, including if Amazon sells allor substantially all or its assets). eBay has also changed its policy to allow for thelimited transfer of its customer information:We have . . . made a policy change to comply with an evolving industrystandard in light of the Federal Trade Commission’s action with theToysmart website. Since that decision, many popular websites have modified

E-tailers also face difficulties in deciding with whom to dobusiness. While the Internet opens markets, it also potentiallysubjects the e-tailer to a range of unknown laws. For example,many countries consider their consumer protection laws to bemandatory. The content of those laws varies by country. Theopen question is whether e-tailers, particularly small businesses,can afford to educate themselves on all of the relevant law orassume the risk of ignorance. Another example is, of course, thecontroversy surrounding a French judge’s order to Yahoo! to blockInternet users in France from accessing United States based sitesthat sell Nazi memorabilia.57Technology may provide a partial solution to some of bothconsumers’ and e-tailers’ concerns. Technology like the Platformfor Privacy Preferences creates a standard for software that cangive Web surfers more information on Websites’ privacy policies aswell as more control over what personal information the viewertransmits to a site.58 Other technology permits firms to maintaingeographical boundaries even on the Internet.59 For example,their privacy policies to state that user information is an asset, and as such,they could sell, transfer, or rent that information to any third party. One ofthe most troubling aspects of such a policy is that it does not clarify how athird party could use your information. Our new policy describes the fewrare instances in which we would transfer your information to a third partyand the limited circumstances in which your information could be used.eBay Frequently Asked Questions about the Privacy Policy Revision, athttp://pages.ebay.com/help/basics/pprevision1-faq.html (last visited Jan. 14, 2002).eBay now also discloses what could happen to its customer information in the eventeBay is acquired or merged with another company:It is possible that eBay, its subsidiaries, its joint ventures, or anycombination of such, could merge with or be acquired by another businessentity. Should such a combination occur, you should expect that eBay wouldshare some or all of your information in order to continue to provide theservice. You will receive notice of such event . . . and we will require that thenew combined entity follow the practices disclosed in this Privacy Policy.eBay Privacy Policy, at http://pages.ebay.com/help/community/png-priv.html (lastvisited Jan. 15, 2002).57. See Mylene Mangalindan & Kevin Delaney, Yahoo! Ordered To Bar the FrenchFrom Nazi Items, Wall St. J., Nov. 21, 2000, available at 2000 WL-WSJ 26617563.58. See W3C Platform for Privacy Preferences Initiative: An Introduction to P3P,at http://www.w3.org/P3P/introduction.html (last visited June 23, 2001) (describinghow the tools following the P3P standard work and how they take action based on theuser’s privacy preferences).59. See Lisa Guernsey, Welcome to the World Wide Web. Passport, Please?, N.Y.Times Online, Mar. 15, 2001, athttp://www.nytimes.com/2001/03/15/technology/15BORD.html (discussing technologythat allows a vendor to determine where a customer is located physically and citing

Universal Music Group uses software to detect its customers’locations to determine whether it can offer a particular customerthe option of a digital download consistent with the copyright lawof the customer’s jurisdiction.60Another solution is global harmonization of the law. Theongoing controversy between European and American conceptionsof privacy demonstrates, however, that this approach may not bepractically workable. Perhaps the best approach would be to seekharmonization on rules governing conflict of laws. Thenjurisdictions would still be able to enforce local values, whileconsumers and e-tailers alike would know what law governs theirtransactions.III. E-COMMERCE IN INFORMATION PRODUCTSThe latter half of the twentieth century saw a move towardincreasing intellectual property protection for information. In theUnited States, Congress enacted laws like the Digital MillenniumCopyright Act61 and the Anti-Cybersquatting ConsumerProtection Act.62 Courts extended patent protection to businessmethods,63 made expansive use of trademark dilution doctrine,64and some have used trespass effectively as a new form ofintellectual property protection.65 At the same time, informationproviders are increasingly using click-wrap contracts andtechnological protection devices to govern access to and use ofinformation. The question is whether intellectual property law, inconjunction with contract and technology, will endow informationproviders with perfect control over their information to thedetriment of the public. Such perfect control could effectivelyMichael Geist, an Internet law expert, as stating, “We are now seeing geographicalzoning online that mirrors geographical zoning offline”); see also Bob Tedeschi, E-Commerce; Borderless is Out; Advertisers Now Want to Know if a Customer Lives inCairo, Egypt or Cairo, Ill., N.Y. Times, Apr. 2, 2001, at C10 (discussing different firmsand their technology).60. See Tedeschi, supra note 59.61. 17 U.S.C. § 1201 (Supp. IV 1998).62. 15 U.S.C. § 1125(d) (Supp. V 1999).63. See, e.g., State St. Bank & Trust Co. v. Signature Fin. Group, Inc., 149 F.3d1368 (Fed. Cir. 1998) (holding computerized accounting system patentable subjectmatter).64. See Mark A. Lemley, The Modern Lanham Act and the Death of CommonSense, 108 Yale L.J. 1687, 1698 (1999) (reviewing court decisions in cybersquattingcases).65. See supra Part II.

eliminate copyright fair use, adversely affect competition bymaking it difficult for second generation creators to access rawmaterial and redistribute wealth without society’s receiving thecorresponding increase in creativity and innovation that theintellectual property laws were designed to foster.It is a bit too early to conclude that employing techniques ofperfect control is a viable business strategy. It is not too early,however, to think critically about how to safeguard the balancebetween creators of works and the public in a digital world. Forexample, policymakers might consider whether approaches otherthan traditional fair use can achieve that doctrine’s goals. In thepatent area, they might consider reforming the internal workingsof the Patent & Trademark Office and the patent re-examinationdoctrine to help ensure that patents issue only on inventionsmeeting statutory standards.Throughout the history of intellectual property protection,society has gone through alternating cycles of over- andunderprotection of creative works. We may now be in a time ofoverprotection. Certainly, technological advances and the onlineenvironment have created a new milieu that merits, at aminimum, an analysis of whether the law has struck theappropriate equilibrium. It is time to rethink the wave of recentlegislation and court decisions, to formulate a sensible test forpreemption of contractual terms, and to consider alternativeapproaches that can safeguard the goals of the intellectualproperty system.CONCLUSIONE-commerce has already become a significant force in theworld economy. Policymakers, however, must face a range ofissues that will affect the nature and growth of e-commerce overthe coming years. Among these issues are access, how to promotepolicies that encourage e-commerce appropriately and how toaddress e-commerce in information products. Certainly, as theUnited States and other countries address these problems, theyshould find guidance not only in their own experience, but also bycomparing approaches to select that which most appropriatelyadvances the relevant policy goals.

Taking UCITA on the Road: WhatLessons Have We Learned?Amelia H. Boss *INTRODUCTIONIn 1999, with a great deal of fanfare, the National Conferenceof Commissioners on Uniform State Laws completed its work on anew statute designed to bring the Conference—and state law—into the new millennium: the Uniform Computer InformationTransactions Act or UCITA. Two years and two state enactmentslater, there are those who argue that UCITA should be the modelnot only for this country, but for other countries—and regionaland international policy makers—to adopt.Should UCITA be an international model? The answer tothat question is both “yes” and “no.” If one speaks of UCITA as aconcept—the concept that there is a need to address a type oftransaction whose use is increasing exponentially in importance intoday’s economy—the transfer of information by contract—thenthe answer is yes. If one views UCITA as standing for theproposition that a comprehensive and accessible body of lawcovering information contracts would add the predictability andcertainty desired by those engaged extensively in electroniccommerce, the answer is yes. And if one views UCITA as a“checklist” of issues that must be confronted in efforts to deal withnew information-based transactions (whether those efforts arethose of a practitioner in drafting a license or a legislatordetermining what issues to address next), the answer is again yes.Unfortunately, however, if one views UCITA as a specific bodyof law, to be enacted substantially as is with little reevaluation,reexamination and reassessment of its provisions (especially in*Charles Klein Professor of Law, Temple University James E. Beasley School of Law.237

the international context), then the answer is no. My involvementin the drafting of UCITA for over fourteen years—from theinception of the idea until almost the end1—has undoubtedlyinfluenced both of my responses to the main question.It has also convinced me that the nature of the debatesurrounding UCITA needs to change. Over the two years sinceUCITA was promulgated by the National Conference ofCommissioners on Uniform State Laws, and indeed dating frombefore its enactment, there has been extensive opposition to theproject.2 Opponents of the legislation have raised concerns overthe adoption process (some concerns meritorious, others merereiterations of concerns without basis or repetition of old concernssince proven wrong), and in the bulk of states to date, thatopposition has led to the failure of UCITA to successfully navigatethe state legislative process or, in some cases, has led toenactment of anti-UCITA legislation.3 In the process, proponentsof the legislation have tried to negotiate compromises, askingopponents to identify those provisions on which there isdisagreement to see whether concerns could be addressed.4 Thus,1. I chaired the Ad Hoc Committee on Scope of the Uniform Commercial Code ofthe American Bar Association’s Section of Business Law, Uniform Commercial CodeCommittee, and in 1985 helped launch the initial study on software contracting underthe UCC. That group ultimately produced a report recommending statutory treatmentof software contracts, at which point I began working with the National Conference ofCommissioners on Uniform State Laws (NCCUSL) to study the potential of this newproject. When the Conference (along with its sister organization, the American LawInstitute) approved the creation of a Drafting Committee on software contracting, Iwas appointed to that committee as the American Law Institute representative andsimilarly served on the Drafting Committee to revise Uniform Commercial Code Article2. When the decision was made to split information/software licensing off from Article2, I continued on the Article 2 Drafting Committee and also served as the AmericanLaw Institute member of the Drafting Committee on Article 2B Licensing. I served inthat capacity until the Spring of 1999, when the American Law Institute withdrewfrom the process, and the project was retitled the “Uniform Computer InformationTransactions Act” (UCITA). At that time, I was asked to continue as an advisor on theproject, but declined to do so. Three months later the National Conference completedits work on UCITA. The complete text of the Act is available athttp://www.law.upenn.edu/bll/ulc/ucita/ucitsFinal00.htm. This article reflectsdevelopments through August 2001.2. See infra notes 18-28 and accompanying text. For a sampling of the depth andbreadth of the opposition, see http://www.4cite.org (last visited Nov. 10, 2001) andhttp://www.badsoftware.com (last visited Nov. 10, 2001).3. See infra note 19 and accompanying text.4. This led to the drafting of amendments to UCITA, but not to any extensiverevisions. The NCCUSL approved amendments to UCITA in 2000 and 2001, makingminor changes to section 103 (excluding motion picture industry), section 216 (adding

the debate has centered on whether to enact UCITA as a whole,enact it with minor amendments, or (at the other extreme) kill thelegislation.The time has come to reevaluate UCITA. It does not makesense to ask what amendments to isolated provisions might makeit “sell” to different constituencies. UCITA contains a number ofcontroversial provisions on which consensus can never beachieved, and the opposition will remain.5 Although it isundoubtedly possible to enact a law about which there is somecontroversy, the fact that UCITA contains so many controversialprovisions compounds the difficulties of enactment. Rather thanfocusing on the controversial provisions and how they can beamended to satisfy the opponents, the focus might better be placedon what positive can be gleaned and saved from UCITA.UCITA stands as a potential roadmap outlining the issues ofimportance for future development, debate and resolution; guidingdevelopments in other venues (such as international law-makingvenues); or identifying those points on which there is sufficientagreement that legislative enactment of those provisions could beachieved. UCITA was (and is) an extremely ambitious project thatinvites and merits intense scrutiny and study. Such a detailedanalysis, however, is beyond the scope of this much more limitedendeavor—to permit someone who was at many stages deeplyinvolved in the process the luxury of ruminating over what hastranspired. I am one who believes that the lessons to be learnedfrom UCITA (and the process by which it was drafted) are worthfar more than its provisions themselves. These lessons may proveextremely valuable to those who toil in the field of legislativereform, both domestically and internationally.idea submission provisions) and section 816 (clarifying limitations on self-help). SeeAmendments to the Uniform Computer Information Transactions Act (ratified Aug. 4,2000), available at http://www.law.upenn.edu/bll/ulc/ucita/ucitaAMD.htm (last visitedOct. 8, 2001); Amendments to Sections 605 and 816 of the Uniform ComputerInformation Transactions Act (as Last Revised in 2000), available athttp://www.ucitaonline.com/docs/0101a.htm (last visited Oct. 8, 2001); AmendmentsApproved by NCCUSL Executive Committee Pending Ratification of the Conference(Feb. 2000), available at http://www.law.upenn.edu/bll/ulc/ucita/approveamend.htm(last visited Oct. 8, 2001).5. An example of this may be § 209, the choice of law provisions. There has beenan inability to reach consensus on them in other fora such as the Hague Conference onPrivate International Law. See Paul Hofheinz, Birth Pangs For Web Treaty SeemEndless, Wall St. J., Aug. 16, 2001, at A11. In UCITA, the real problem comes from thebundling together of many controversial issues such as this into one package.

I. A BIT OF BACKGROUNDOver the last ten years, our economic marketplace hasdramatically changed in many significant ways.6 With the adventof the Internet and the rise of computer technologies generally, wehave witnessed the commoditization of information: informationitself has become the subject of commercial transactions, not justthe medium of performing them. This is, in essence, a key featureof our emerging information economy. Business can now beconducted at lightning speed between parties at great distancesfrom one another; national boundaries and cultural differences areboth invisible. There have been increasing demands on the law,developed in the context of different transactions, to catch up andadapt to these new transactions; cries for clarity, certainty, andinternational uniformity abound.7 Uniformity eliminates the“jurisdictional risk” of non-uniform law in cross-border activities,which is particularly a problem in Internet transactions. Acontractual code, it is argued, will facilitate information exchange;codification lowers negotiation costs by supplying off-the-shelfterms parties can adopt, filling gaps if they cannot agree andproviding the backdrop against which negotiations occur.86. See, e.g., Neil B. Cohen, Introduction to Symposium, Software as aCommodity: International Licensing of Intellectual Property, 26 Brook. J. Int’l L. 3(2000). For an interesting analysis of how technology has transformed society over alonger period, see M. Ethan Katsh, The Electronic Media and the Transformation ofLaw (1989).7. See, e.g., William J. Clinton & Albert Gore, Jr., A Framework for GlobalElectronic Commerce (1997), available at http://www.iitf.nist.gov/eleccomm/ecomm.htm(last visited Nov. 11, 2001) [hereinafter Clinton & Gore, Framework].Many businesses and consumers are still wary of conducting extensivebusiness over the Internet because of the lack of a predictable legalenvironment governing transactions. This is particularly true forinternational commercial activity where concerns about enforcement ofcontracts, liability, intellectual property protection, privacy, security andother matters have caused businesses and consumers to be cautious.Id. The White House report noted the work being done by the United Nations and theNational Conference of Commissioners on Uniform State Laws in developing rulesgoverning electronic documents and signatures, and observed:The United States Government supports the adoption of principles alongthese lines by all nations as a start to defining an international set of uniformcommercial principles for electronic commerce. We urge UNCITRAL, otherappropriate international bodies, bar associations, and other private sectorgroups to continue their work in this area.Id. (emphasis added).8. See, e.g., Mary Jo Howard Dively & Carlyle C. Ring, Jr., Overview of Uniform

It is against this background that UCITA was drafted. Thedrafting history of UCITA has been described at lengthelsewhere.9 Nonetheless, a few critical aspects of that historybear repeating.First, the concept of statutory treatment of softwaretransactions originally surfaced over fifteen years ago ascommercial lawyers grappled with judicial attempts to apply thegoods provisions of the Uniform Commercial Code to software.10At the outset, the issues and the discourse were framed from acommercial perspective, focusing on contracting issues such aswarranties and remedies in the software context; only later in theprocess (as more and more software and information industryrepresentatives became involved) did the focus shift from thepotential of a new software contracting statute to more expansivetreatment of information and additional protections for licensorsof information.11 This evolution—from contract issues toComputer Information Transactions Act, in Advanced Licensing Agreements for theNew Economy 2001, 201 (P.L.I. ed., 2001); Jeff C. Dodd, Time and Assent in theFormation of Information Contracts: The Mischief of Applying Article 2 to InformationContracts, 36 Hous. L. Rev. 195 (1999); Philip M. Nichols, Electronic Uncertaintywithin the International Trade Regime, 15 Am. U. Int’l L. Rev. 1379 (2000).9. There has been an exceptional amount of literature devoted to UCITA. Atleast two law review symposia issues have dealt with the topic. Symposium, Licensingin the Digital Age, 36 Hous. L. Rev. 1 (1999); Symposium, Approaching E-CommerceThrough Uniform Legislation: Understanding the Uniform Computer InformationTransactions Act and the Uniform Electronic Transactions Act, 38 Duq. L. Rev. 205(Winter 2000). See, e.g., Raymond T. Nimmer, Images and Contract Law—What LawApplies to Transactions in Information, 36 Hous. L. Rev. 1 (1999) [hereinafter Nimmer,Images]; Raymond T. Nimmer, Through the Looking Glass: What Courts and UCITASay about the Scope of Contract Law in the Information Age, 38 Duq. L. Rev. 255 (2000)[hereinafter Nimmer, Looking Glass]; Raymond T. Nimmer & Holly K. Towle, UCCArticle 2B Would Govern Software Licensing, Nat’l L.J., Feb. 12, 1996, at C2; Holly K.Towle, Mass Market Transactions in the Uniform Computer Information TransactionsAct, 38 Duq. L. Rev. 371 (2000); Holly K. Towle, The Politics of Licensing Law, 36 Hous.L. Rev. 121 (1999).10. The concept of a statute devoted to software arose out of the work of an AdHoc Subcommittee on the Scope of the Uniform Commercial Code Committee of theAmerican Bar Association and was first discussed in print in 1988. See Amelia H. Boss& William J. Woodward, Jr., Scope of the Uniform Commercial Code: Survey ofComputer Contracting Cases, 43 Bus. Law. 1513 (1988); Amelia H. Boss et al., Scope ofthe Uniform Commercial Code; Advances in Technology and Survey of ComputerContracting Cases, 44 Bus. Law. 1671 (1989); Jeffrey B. Ritter, Scope of The UniformCommercial Code: Computer Contracting Cases and Electronic Commercial Practices,45 Bus. Law. 2533 (1990).11. This shift occurred during the period when the use of licensing to govern andrestrict the use of information gained importance and visibility as courts and Congressrefused to grant copyright protection to information such as databases. The prominence

information policy issues—increased the visibility of the product.Moreover, it significantly added to the complexity of the projectand the number of economic interests affected by the proposal,making consensus on such a statute harder to achieve andcontributing to the growing dissatisfaction with its provisions.Second, the efforts to address software contracts were initiallycombined with other efforts to revise domestic sales law—arecognition that in the generic area of contracting, there was agreat deal of overlap between contracts for the transfer of goodsand contracts for the transfer of information.12 Although it wasacknowledged that certain aspects of information contracts mightrequire different provisions, the similarities were deemedsufficient enough, at the outset, to justify a core set of provisions(“hub” principles) governing both goods and information contracts,with special rules (“spoke” rules) necessary to deal with theunique aspects of each.13 This recognition (that there areof licensing also illustrated the desire of licensors to place additional restrictions on theuse of copyrighted information not otherwise granted by copyright law. See David A.Rice, Legal-Technological Regulation of Information Access, in Libraries, Museums,and Archives: Legal Issues and Ethical Challenges in the New Information Era 275(Tomas A. Lipinski, ed. 2002). A second significant development occurred during thisperiod that profoundly affected the substance of the UCITA discussions. Althoughearly cases, such as Vault Corp. v. Software Ltd., 847 F.2d 255 (5th Cir. 1988) andStep-Saver Data Sys. v. Wyse Tech., 939 F.2d 91 (3d Cir. 1991), cast significant doubton the enforceability of licenses in the software context, two significant cases from theSeventh Circuit Court of Appeals, ProCD v. Zeidenberg, 86 F.3d 1447 (7th Cir.1996),and Hill v. Gateway 2000, Inc., 105 F.3d 1147 (7th Cir.1997), cert. denied, 118 S. Ct. 47(1997), upheld this method of contracting. This development permitted the softwareindustry, in particular, to refocus its efforts on what it hoped to achieve from theUCITA drafting process.12. From 1991 to 1995, the Drafting Committee to revise Article 2 considered notonly the substance of the goods provisions, but also the question of whether to dealwith software in the context of Article 2—in what was dubbed as a “hub-and-spoke”approach—or in a separate statute (whether or not a part of the UCC). See Linda J.Rusch, A History and Perspective of Revised Article 2: The Never Ending Saga of aSearch for Balance, 52 SMU L. Rev. 1683, 1715 (1999); Richard E. Speidel,Introduction to Symposium on Proposed Revised Article 2, 54 SMU L. Rev. 787 (2001)[hereinafter Speidel, Symposium Intro]; Richard E. Speidel, Revising UCC Article 2: AView From the Trenches, 52 Hastings L.J. 607 (2001) [hereinafter Speidel, Trenches].Folding the discussions of how to treat software licensing into the ongoing Article 2sale of goods revision process was an implicit recognition of the substantial overlapbetween the substantive rules governing each type of transaction.13. Under the proposed “hub-and-spoke” approach, the core contracting principleswould apply to every transaction: sales (existing Article 2); leases (existing Article 2A)and software licenses. Potentially, they could also apply to other types of contracts,such as services contracts. These core principles would be followed by “spokes” settingforth specific rules for the particular transaction at hand.

similarities as well as differences between information and goodscontracts) is frequently overlooked in the literature proclaimingthe need for an entirely separate body of information law.14Third, the critical decision to separate goods from softwareand information during the drafting process was not, as somehave portrayed, the result of a determination that contract issuessurrounding the licensing of information were so distinct fromthose involving the sale of goods that a different body of ruleswere required. By 1995, the drafting committee originallycharged with responsibility in the field was prepared torecommend a “hub-and-spoke” treatment for goods andinformation. A central core of basic contracting principlesgoverning all transactions (sales, leases, licenses and possiblyother transactions such as services) would be combined withspecific rules where needed. Nonetheless, in 1995 the leadershipof the National Conference of Commissioners on Uniform StateLaws made the surprising announcement that it was abandoningattempts to treat software within Article 2, and was insteadcreating a new Drafting Committee (the Article 2B DraftingCommittee) with a new reporter to draft a new statute on thelicensing of information.15 This split was not justified on thegrounds that information and software were significantlydifferent, but rather on the basis that the logistics of restructuringArticle 2 into a hub with spokes would require extraordinary timeand resources.My perception is that an additional, unarticulated motive wassurely behind this surprising decision. As long as sales andsoftware were being handled by the same Drafting Committee,14. See, e.g., Nimmer, Looking Glass, supra note 9. Professor Nimmer examinesthe cases struggling with which law (Article 2 or common law) to apply to informationtransactions, but often fails to examine why the question is important. Asking whetherthe licensing of information is the same as, or different from, the sale of goods is thesame as asking “is the glass half empty or half full.” The answer is “it depends.” If oneis in need of a drink, the salient fact is that there is something in the glass (it is halffull); if the question is whether to refill the glass, the salient fact is that it is halfempty. In the context of UCITA, the question is not whether licensing of information isdifferent than sales of goods, but whether those differences in the transactions justify adifference in the applicable legal rule.15. It is noteworthy that this important step was taken by the leadership of theNCCUSL with minimal input from its partner in the Code revision process, theAmerican Law Institute, and with minimal input from the Chair, Reporter or DraftingCommittee on Article 2. For a diplomatic account of the failure of the hub-and-spokeapproach, see Speidel, Trenches, supra note 12, at 612-15.

decisions on fundamental issues—such as the treatment ofadhesion contracts, the nature of assent, the role of formalities incontract, the availability of damages—would be the same. Yetpressure was being placed on the Conference by the softwareindustry, which primarily represents the interests of licensors andwas uncomfortable with what it perceived as the liberal “probuyer”or “pro-consumer” positions being advanced in the contextof Article 2.16 Thus, splitting off the two committees allowed forthe development of the licensing law in a different direction thanthe sales article. In attempts following the creation of the Article2B Drafting Committee to coordinate the two drafts, both thedegree of divergence between the two and their fundamentallydifferent approaches to contract law became apparent.17Fourth, UCITA in the United States has not met with totalacceptance—it has been enacted in only two states to date.Indeed, one might characterize its reception in some-quarters asdown right hostile. For example, the attorneys general in manystates opposed the original adoption of UCITA by the NationalConference.18 In addition to lobbying against enactment ofUCITA by states, opponents have actually gone on the offensiveand urged state legislatures to adopt legislation that wouldaffirmatively preclude application of UCITA by courts in that stateeven where the parties have chosen UCITA as the governing bodyof law.19 Opposition and dissension has surfaced in other bodies16. Id.17. One of the key results of the split is that the Article 2 Drafting Committee wasunder increased pressure—in the name of “uniformity”—to revise its provisions tofollow the more conservative ones in Article 2B. This impact was felt in such areas asthe Statute of Frauds (which the Article 2 drafting committee originally voted toeliminate but which was vigorously defended by the Article 2B committee; it was laterreinstated in Article 2), the parol evidence rule (Article 2 wanted to liberalize the rule,but Article 2B strengthened it; ultimately, Article 2 was made to conform to Article2B), and contract policing (Article 2 contained additional provisions absent in Article2B that were eventually eliminated).18. Letter from State Attorneys General to Gene N. Lebrun, President of theNCCUSL (July 23, 1999), available at http://www.2bguide.com/docs/799ags.html (lastvisited Sept. 8, 2001); Letter from Attorneys General of 11 States to Gene Lebrun,President of the NCCUSL (July 28, 1999), available athttp://www.2bguide.com/docs/799mags.html (last visited Sept. 8, 2001) (agreeing withJuly 23, 1999, letter).19. Three states have already adopted such “anti-UCITA” legislation. Ironically,these states (Iowa, West Virginia and North Carolina) have included such a provisionin their enactment of the Uniform Electronic Transactions Act, another product of theNational Conference. See Uniform Electronic Transactions Act, Iowa Code Ann. §

as well, such as the American Bar Association, which hadobservers attending all drafting committee meetings and that isnormally called upon to endorse uniform acts produced by theNational Conference.20 The academic literature abounds withcritiques of UCITA21 and even regulatory bodies such as theFederal Trade Commission have expressed reservations about the554D (West Supp. 2001), 2000 Iowa Legis. Serv. H.D. 2205 (West) (last visited Aug. 8,2001), available athttp://www.legis.state.ia.us/GA/78GA/Legislation/HF/02200/HF02205/Current.html(codified at Iowa Code § 554D.104 (repealed 2001)) (“A choice of law provision . . . whichprovides that the contract is to be interpreted pursuant to the laws of a state that hasenacted the uniform computer information transactions Act . . . or any substantiallysimilar law, is voidable.”); Uniform Electronic Transactions Act, W. Va. Code § 55-8-15(2001), 2001 W. Va. Leg. Serv. 120 (West); Uniform Electronic Transactions Act, ch.2001 N.C. Sess. Laws 2001-295 sec. 66-329 (July 21, 2001). Ironically, at the time WestVirginia passed this provision, it was the home of the president of the NCCUSL. Morerecently, the New York Attorney General’s office proposed legislation would declarethat UCITA violates New York public policy. The concern of the N.Y. Attorney Generalwas the impact of UCITA on consumers, and particularly UCITA’s validation of clickwrapcontracting practices and licenses that “diminish significant rights andprotections established over many years for the protection of consumers” in New York.N.Y. Attorney General’s Legislative Program Bill No. 33-01, 12 BNA Electronic Com. &L. Rep. 288 (2001); see Am. Online, Inc. v. Superior Court, 108 Cal. Rptr. 2d 699 (Cal.Ct. App. 2001) (invalidating a choice of forum and law clause invoking Virginia law,including UCITA, as contrary to California public policy.).20. Typically, upon the completion of new proposed uniform legislation, theNational Conference places the item on the agenda of the House of Delegates of theAmerican Bar Association for ratification by that group. In the year following thecompletion of UCITA, the National Conference submitted a resolution calling for ABAapproval of another piece of electronic commerce legislation, the Uniform ElectronicTransactions Act, but failed to put UCITA on the agenda of the House. To the extentthat this failure illustrated a perception that there would be a fight in the House overany such resolution, the perception has proven to be true. In the Summer of 2001, aresolution was introduced in the House that would disapprove of UCITA and call uponthe Conference to withdraw it as a proposed statute for state enactment. Thisresolution was withdrawn by its sponsor, the Torts Insurance and Practice Section,pending additional discussions between the ABA and the National Conference onsubstantive objections to UCITA. In January 2002 an ABA Working Group issued itsreport on UCITA, describing it as “a very complex statute for even knowledgeablelawyers to understand and apply,” and concluding that in addition to requiringsubstantial changes in many of its sections, UCITA should be “redrafted to make iteasier to understand and use.” American Bar Association Working Group Report onthe Uniform Computer Information Transactions Act (UCITA), Jan. 31, 2002, availableat http://www.abanet.org/leadership/ucita.pdf.21. See, e.g., Symposium, Intellectual Property and Contract Law in theInformation Age: The Impact of Article 2B of the Uniform Commercial Code on theFuture of Information and Commerce, 13 Berkeley Tech. L.J. 809 (1998) (criticizing thepredecessor of UCITA); Symposium, Intellectual Property and Contract Law for theInformation Age: The Impact of Article 2B of the Uniform Commercial Code on theFuture of Information and Commerce, 87 Cal. L. Rev. 1 (1999) (same).

product.22 Whether or not one agrees with the arguments raisedby opponents of UCITA, the fact remains that at this stage thereis no consensus about its acceptability.Last, events occurring in 1999 have undoubtedly had the mostsignificant impact on Article 2B (soon to become UCITA) and itschances of enactability: attempts to include the treatment ofinformation within the Uniform Commercial Code (as Article 2BLicensing) were abandoned; the American Law Institute withdrewfrom the process; and the National Conference reformulated thedraft as a freestanding uniform act. The only “official” reasongiven for the split, according to the joint press release of the twoorganizations, was that “this area [computer informationtransactions] does not presently allow the sort of codification thatis represented by the Uniform Commercial Code.”23 The problemwas much more fundamental.The decision to part ways with Article 2B came after severalyears of mounting tension about this project between the twosponsoring organizations.24 In 1998, and again in 1999, the22. In October of 2000, the Federal Trade Commission held a Public Workshop onWarranty Protection for High-Tech Products and Services devoted almost exclusively towhether the FTC should enact consumer protection provisions to supplement oroverride UCITA. See Federal Trade Commission: Warranty Protection for High-TechProducts and Services, at http://www.ftc.gov/bcp/workshops/warranty/ (last visitedSept. 8, 2001); Letter from the Bureau of Consumer Protection, Bureau of Competition,Policy Planning of the Federal Trade Commission, to Carlyle C. Ring, Jr., Chairman ofthe NCCUSL Article 2B Drafting Committee (Oct. 30, 1998), available athttp://www.ftc.gov/be/v980032.htm (last visited Sept. 8, 2001) (expressing concernsabout Article 2B); Letter from the Bureau of Consumer Protection, Bureau ofCompetition, Policy Planning of the Federal Trade Commission, to John L.McClaugherty, Chair of the Executive Committee of the NCCUSL (July 9, 1999),available at http://www.ftc.gov/be/v990010.htm (last visited Sept. 8, 2001) (stating thatconcerns of prior letter were not addressed in any significant respect).23. Joint Press Release, ALI-NCCUSL, NCCUSL to Promulgate FreestandingUniform Computer Information Transactions Act—ALI and NCCUSL Announce thatLegal Rules for Computer Information Will Not Be Part of UCC (Apr. 7, 1999),available at http://www.law.upenn.edu/bll/ulc/ucita/2brel.htm (last visited Dec. 18,2001) (“The Conference believes that UCITA can provide a framework in which soundbusiness practices may further evolve in the marketplace bounded by standards ofappropriate public policy.”). The press release does not say that the American LawInstitute shared that belief.24. It should be noted that the “tension” referred to is not tension between the twobodies as a whole, but a tension between the leadership of those bodies. UCITA wasnot without controversy even within the NCCUSL process. Objections to UCITA wereso strong that the 1999 motion to refer UCITA for a final vote by the Conference(normally a pro forma matter) was highly debated and only passed by a 37-11 vote withfive abstaining; the final vote of the states on UCITA was 43-6, with two states voting

American Law Institute refused to put Article 2B on the agendafor approval at its annual meetings.25 Even in its stance as a“discussion draft,” Article 2B attracted much criticism at theannual ALI meetings, precipitating significant motions forchanges in substance.26 An ALI official publication, reporting onthe joint decision to remove Article 2B from the UniformCommercial Code, explained that while the National Conferencebelieved that “the proposed statute is currently ready to provide aviable legal framework for the evolution of sound businesspractices in computer-information transactions,” the AmericanLaw Institute Council “continued to have significant reservationsabout both some of its key substantive provisions and its overallclarity and coherence.”27 The substantive basis for the ALIto abstain and two states not voting at all. [The total number of states is fifty-three notfifty, as the District of Columbia, Puerto Rico and the Virgin Islands are also counted.]The final approving vote was undoubtedly influenced by pleas from Conferenceleadership who maintained that unless UCITA was passed, the Conference would bepreempted by Congress (an event which did not come to pass) and who asked thatUCITA be passed to allow a field test of its provisions in the states with no obligationon individual commissioners to support it. Many undoubtedly expected UCITA to die anatural death upon its final promulgation.25. See Press Release, ALI, Major Projects Scheduled for Completion at AmericanLaw Institute’s 75th Annual Meeting (May 1, 1998), available athttp://www.ali.org/ali/pr0501.htm (last visited Sept. 8, 2001) (“[A] special ad hoccommittee of the Institute’s Council has recommended that ALI take no final actionthis year because of present concerns about both the architecture and scope of 2B.”);Institute to Review Seven Drafts at 1999 Annual Meeting, A.L.I. Rep. (Winter 1999),available at http://www.ali.org/ali/Rptr_7drafts.htm (last visited Sept. 8, 2001) (“Whilethe leadership of NCCUSL also plans to seek final approval of Article 2B this year, theALI Council continues to have significant reservations about 2B, especially concerningits provisions on scope, invalidation for fundamental public policy, and assent to posttransactionterms, but also regarding its overall coherence and clarity.”).26. See, e.g., Jean Braucher & Peter Linzer, Motion Regarding Assent Issues inProposed UCC Article 2B (May 5, 1998), at http://www.ali.org/ali/Braucher.htm (lastvisited Nov. 17, 2001); Charles R. McManis, Motion Regarding UCC Article 2B § 2B-308 (Tentative Draft Apr. 14, 1997) (May 9, 1997), available athttp://www.ali.org/ali/McManis.htm (last visited Nov. 17, 2001); Charles R. McManis,Motion Regarding UCC Article 2B § 2B-208 (Mass Market Licenses) and § 2B-105(Relationship to federal law)(Tentative Draft, Apr. 15, 1998) (May 5, 1998), available athttp://www.ali-aba.org/ali/McManis2.htm (last visited Sept. 8, 2001); William J.Woodward, Jr., Motion to Delete Section 2B-107(a) from Draft UCC Article 2B (May 6,1998), at http://www.ali.org/ali/Woodward1.htm (last visited Nov. 17, 2001); William J.Woodward, Jr., Motion to Delete Section 2B-108 from Draft UCC Article 2B (May 6,1998), at http://www.ali.org/ali/Woodward2.htm (last visited Nov. 17, 2001). Foradditional motions, see http://www.ali.org/ali/Motions.htm (last visited Nov. 17, 2001).27. Article 2B Is Withdrawn from UCC and Will Be Promulgated by NCCUSL asSeparate Act, A.L.I. Rep. (Spring 1999), available at

decision could not be clearer. Following the withdrawal of the ALIfrom the process, the three ALI members of the draftingcommittee were asked to remain as advisors, but declined, citing“a number of underlying concerns including matters of substance,process, and product:”In terms of product, the draft has, in attempting toaddress numerous concerns of affected constituencies,progressively moved away from articulating sufficientand generally applicable default rules towardestablishing increasingly particular and detailed rules.In so doing, the draft sacrificed the flexibility necessary toaccommodate continuing fast-paced changes intechnology, distribution, and contracting. In terms ofprocess, the guiding principle appeared to be theConference’s desire to expedite approval and commencehttp://www.ali.org/ali/R2103_Art2b.htm (last visited Sept. 8, 2001); see also Geoffrey C.Hazard, Jr., Report of the Director, Address Before the American Law Institute (May1999), in 76 A.L.I. Proc. 11 (2000), available at http://www.ali.org/AR99_director.htm(last visited Sept. 8, 2001) (“There has been great difficulty in arriving at a suitabledefinition of the scope of the provisions aimed at licensing and some differencesconcerning the provisions on electronic contracting.”); Letter from Geoffrey C. Hazard,Jr., Director of the ALI, to Gene N. Lebrun, President of the NCCUSL, and Charles A.Wright, President of the ALI (Mar. 26, 1998), athttp://www.2bguide.com/docs/ghmar98.html (last visited Sept. 8, 2001) (stating thatthe text “needs significant revision” and expressing concerns regarding scope,architecture, clarity of expression, relation to other law, electronic contracting, andother controversial issues); Memorandum from Geoffrey C. Hazard, Jr., et. al., toDrafting Committee on Uniform Commercial Code Article 2B-Licenses, July 1998Draft: Suggested Changes (Oct. 9, 1998), available athttp://www.2bguide.com/docs/gch1098.pdf (making suggestions: scope too broad; needto cover all access contracts questionable; post-contract assent provisions not trueassent; contract formation provisions should be drafted by an appropriate group “thatconsiders these matters in broader context” than Article 2B); Memorandum from DavidA. Rice, ALI Drafting Committee Member, to Article 2B Drafting Committee (Mar. 18,1998), available at http://www.2bguide.com/docs/ricemar.html (last visited Sept. 8,2001) (commenting and proposing changes to Article 2B); Memorandum from David A.Rice to ALI Council Subcommittee on Article 2B, Critical Considerations Concerningthe Readiness of Article 2B (Mar. 17, 1998), available athttp://www.2bguide.com/docs/ricemar.html#alic (last visited Nov. 17, 2001);Memorandum from Amelia H. Boss, et al., to UCC2B Drafting Committee andRaymond T. Nimmer (Nov. 4, 1996), available at www.2bguide.com/docs/polmem.html(last visited Setp. 8, 2001) (proposing policy issues for resolution). Some took greatoffense at the ALI’s criticism of Article 2B, reducing the debate to (anonymous) adhominem attacks. See A Firefly on the Wall: The UCC2B Experience, available athttp://www.2bguide.com/docs/afotw.html (last visited Sept. 8, 2001) (responding toGeoffrey Hazard’s comments made at the meeting).

enactment of the draft. This led to obviating rather thanlearning from strong concerns expressed by Conferenceand Institute discussions over the entire course of theproject, ranging from scope and drafting to the interplaywith intellectual property rules. Substantively . . . thethree of us often disagree. Yet we believe that somerules, although they may assure importantconstituencies’ support for the draft, nonethelessjeopardize enactability because of the ultimate balance ofinterests achieved.These are not new, or newly expressed, concerns. Theyare fundamental concerns and have been aired before inConference and Institute discussions, by individualmembers, Drafting Committee members and observers,and Internet discussion list participants, as well as bysoftware and other computer science enterprises andprofessional organizations, law professors, and editorialwriters. The persistent din of these concerns hascontributed significantly to our decision to decline theinvitation to participate as advisors.28II. THE INTERNATIONAL DEBATEA. Goods, Services, or Sui Generis: Herein the application of theCISGAlthough UCITA is a domestic statute written against thebackdrop of US law, the pressures that were behind its evolutionare present on the international level as well. With the growth ofinformation transactions, there has been increasing recognitionthat many of the primary sources of law governing internationalcommercial transactions do not effectively cover information. Justas the UCITA discussions had their origins in dialogue overwhether and to what extent Article 2 of the Uniform Commercial28. See Memorandum from David Bartlett et al., to Uniform ComputerTransactions Act Drafting Committee (May 7, 1999), available athttp://www.2bguide.com/docs/50799dad.html (last visited Sept. 8, 2001). It should beemphasized that the three of us were seldom aligned on individual substantive issuesthat came before the drafting committee; nonetheless, we were united in our decisionsto recommend that the ALI withdraw from the project and to decline to continue ourown participation.

Code might be adapted to cover information,29 the same questionis being asked globally (in case law, in academic literature and inthe legislative process) about the UNCITRAL Convention on theInternational Sale of Goods (CISG).30 Should the internationalequivalent of UCC Article 2, the United Nations Convention onthe International Sale of Goods, be modified to accommodate nongoodstransactions?31The scope of the CISG is limited, both in terms of the subjectmatter of the transaction (goods) and the structure of thetransaction (sales).32 Although the Convention is limited to goods,the term is not defined.33 Thus, there is the question of whetherthe term “goods” should be construed to apply to “virtual goods.”The argument has been made by legal commentators that the saleof software may fall under the Convention’s substantive sphere ofapplication, even though it is not tangible, with the qualificationbeing that the software be “off the rack” software i.e., not custommadesoftware or software that has been extensively modified tomeet the user’s needs.34 Recent judicial decisions have supported29. See Boss & Woodward, supra note 10; Boss et. al., supra note 10; Ritter, supranote 10.30. United Nations Convention on Contracts for the International Sale of Goods,art. 19(3), U.N. Doc. A/CF.97/18 (1980), available athttp://www.uncitral.org/english/texts/sales/CISG.htm (last visited Sept. 15, 2001)[hereinafter CISG]. For example, the question of whether to classify softwaretransactions as goods transactions arises in contexts other than the sale of goods, suchas their treatment under trade agreements within the WTO. See, e.g., Stewart A.Baker et al., E-Products and the WTO, 35 Int’l Law. 5 (2001); Joseph Tasker, Jr., TheInformation Technology Agreement: Building A Global Information InfrastructureWhile Avoiding Customs Classification Disputes, 26 Brook. J. Int’l L. 917, 947 (2001).31. For selected articles on the extension of the Convention on the InternationalSale of Goods to software see infra note 34. The question of whether informationtransactions are analogous to goods transactions is being debated in many areas inaddition to the area of general contract law. In WTO discussions, for example, thequestion is whether information fits within GATT (goods) or GATS (services).There is a second issue raised about the coverage of the CISG: whether itshould deal with electronic transactions—i.e., those contracts electronically created.The question of the medium of the communication, dealt with by the UNCITRALModel Law on Electronic Commerce, is distinct from the question of the nature of thesubject matter of the transaction.32. CISG, supra note 30, at Annex I. A third limitation in the application of theCISG is that it is inapplicable to consumer transactions, which currently represent asignificant proportion of international electronic commerce.33. Id. 21-22.34. See, e.g., Christina H. Ramberg, The E-Commerce Directive and Formation ofContract in a Comparative Perspective, 26 Eur. L. Rev. 449 (2001) (urging extension ofCISG to services and intangibles to deal with electronic commerce and noting that the

such application.35 To the extent, however, that the result isarguably unclear, it has been maintained that the CISG should beextended to services and intangibles to deal with electroniccommerce.36The United Nations Commission on International Trade Law,which was responsible for the drafting of the CISG, has also beena key player in the development of legal structures for electroniccommerce: in 1996, it completed its work on the Model Law onElectronic Commerce,37 and in 2001 it completed work on theModel Law on Electronic Signatures.38 In its deliberations in thewinter of 2001 on potential future work in the area, theUNCITRAL Working Group on Electronic Commerce turned itsattention to the CISG.39 In addition to looking at whether thePrinciples of European Contract Law and the UNIDROIT Principles on InternationalCommercial Contracts had already extended many of the provisions of the CISGoutside the goods context); Frank Diedrich, Maintaining Uniformity in InternationalUniform Law Via Autonomous Interpretations: Software Contracts and the CISG, 8Pace Int’l L. Rev. 303, 336 (1996) (stating that CISG can be applied to computersoftware); Trevor Cox, Chaos Versus Uniformity: The Divergent Views of Software inthe International Community, 4 Vindabona J. Int’l Com. L. & Arb. 3 (2000); see alsoJames E. Bailey, Facing the Truth: Seeing the Convention on Contracts for theInternational Sale of Goods as an Obstacle to a Uniform Law of International Sales, 32Cornell Int’l L.J. 273 (1999) (challenging belief that CISG reaches uniformity); MarcusG. Larson, Applying Uniform Sales Law to International Software Transactions: TheUse of the CISG, Its Shortcomings, and a Comparative Look at How the Proposed UCCArticle 2B Would Remedy Them, 5 Tul. J. Int’l & Comp. L. 445 (1997) (discussing theapplicability of CISG to software sales).35. OLG Koln, 26 August 1994, Neue Juristische Wochershrift Rechtsprechungs-Report 246 (1995) = CLOUT case n. 122; OLG Koblenz, 17 September 1993, Recht derinternationale Wirtschaft 934 (1993) = CLOUT case n. 281. The CLOUT reference is tothe UNCITRAL project (Case Law on UNCITRAL Texts) reporting on all casesinvolving the interpretation of the CISG and other UNCITRAL products. Seehttp://www.uncitral.org.36. See supra note 34.37. U.N. Comm’n on Int’l Trade Law, UNCITRAL Model Law on ElectronicCommerce with Guide to Enactment (Dec. 16, 1996), available athttp://www.uncitral.org/english/texts/electcom/ml-ecomm.htm (last visited Sept. 8,2001).38. U.N. Comm’n on Int’l Trade Law, UNCITRAL Model Law on ElectronicSignatures, U.N. Doc. A/56/17, Supp. 17, Annex II (June 23, 2001), available athttp://www.uncitral.org/english/texts/electcom/ml-elecsig-e.pdf (last visited Nov. 17,2001); see also Guide to Enactment of the UNCITRAL Model Law on ElectronicSignatures, U.N. Doc. A/CN.9/WG.IV/WP.88 (Jan. 30, 2001), available athttp://www.uncitral.org/english/workinggroups/wg_ec/wp-88e.pdf (last visited Sept. 8,2001).39. See Legal Aspects of Electronic Commerce: Possible Future Work in the Field ofElectronic Contracting: An Analysis of the United Nations Convention on Contracts for

sales Convention needs to be updated to take into account theelectronic formation of contracts, the question was raised as towhether its scope should be extended beyond goods toinformation—or what the report called “virtual goods.” “It isapparent that a clarification of whether the software should beconsidered as ‘goods’ in the sense of the Convention would beuseful in order to ensure uniformity.”40 Revisions could clarifythat the CISG covers all software, only software incorporated intotangible goods, or only off-the-shelf (i.e., non-custom) software.41The scope issue under the Convention involves not only theissue of what constitutes “goods” but what constitutes a “sale.”Much software is not sold but rather licensed. “The differences inthese approaches are considerable. A sales contract, for instance,frees the buyer (i.e., “user”) from restrictions as to the use of theproduct bought and, thus, clearly delineates the boundaries ofcontrol that may be exercised . . . In contrast, a license agreementallows the producer or developer of ‘virtual goods’ (or services) toexercise control over the product down through the licensingchain.”42The issues confronting UNCITRAL then included whether thescope of the CISG should be broadened to pick up softwarelicensing agreements, or whether the rules derived from the CISGshould be developed for these kinds of transactions. The WorkingGroup report to UNCITRAL, however, placed more emphasis onthe application of the CISG to contracts concluded by electronicmeans for the international sale of tangible goods “in view of theurgent need for the introduction of the legal rules that would beneeded to bring certainty and predictability to the internationalregime governing Internet-based and other electronic commercetransactions.”43 Apparently recognizing the controversysurrounding the enactment of UCITA, the report was cautious inrecommending any expansion of the CISG to non-tangible goodstransactions: “Broadening the scope of such work so as to includethe International Sale of Goods, U.N. Doc. A/CN.9/WG/IV/WP.91 (Feb. 9, 2001)[hereinafter Legal Aspects].40. Id. 24.41. Id.42. Id. 28.43. U.N. Comm’n on Int’l Trade Law, Report of the Working Group on ElectronicCommerce on the Work of Its 38th Session, U.N. Doc. A/CN.9/484, 95 (Apr. 14, 2001).The issues preliminarily identified were: offer and acceptance; receipt and dispatch;internationality of the sales transaction; and parties to the sales transaction.

transactions involving goods other than tangible goods, such asthe so-called “virtual goods” or rights in data, was an avenue thatshould be approached with caution, given the uncertainty ofachieving consensus on a harmonized regime.”44 Nonetheless, thereport cryptically noted:There was general agreement within the WorkingGroup that existing international instruments, notablythe United Nations Sales Convention, did not cover avariety of transactions currently made online and that itmight be useful to develop harmonized rules to governinternational transactions other than sales of movabletangible goods in the traditional sense. The WorkingGroup proceeded to consider what elements should betaken into account to define the scope of application ofsuch a new international regime.45Ultimately, the Working Group to the Commissionrecommended that preparation of a legal instrument “dealing withcertain legal issues in electronic contracting be begun on a prioritybasis.”46 What is unclear is the extent to which that instrumentwill include work in the area of software, computer informationand licensing.47 Also unclear is the extent to which UCITA will beused as a model in those deliberations. The Secretariat, in itsreport to the Working Group on the issue, noted the existence ofother instruments, including UCITA, designed to harmonizecertain areas of electronic commerce and suggested that furtherstudy of these rules needed to be undertaken.48 Undoubtedly44. Id.45. Id. 115 (emphasis added).46. Id. 134.47. There are “electronic contracting” and scope issues under the CISG that do notimplicate the question of whether the CISG should be extended beyond the sale ofgoods: how to determine the “internationality” of a transaction for purposes ofdetermining the applicability of the Convention; the “place of conclusion of thecontract”; who are parties to the sales convention, especially when automated systemsor “electronic agents” are used; coverage of consumer contracts and particularlyconsumer protection issues; offer and acceptance, including manifestation of assent;receipt and dispatch. Id. 96-113, 119-22, 125-27.48. Legal Aspects, supra note 39, 7.As an example of such rules that may require further study, the UniformComputer Information Transactions Act (UCITA) was developed in theUnited States of America, since it was felt that the approach of the ‘sale ofgoods’ transactions embodied in the Uniform Commercial Code (UCC) was

ecause of a recognition of the controversial nature of UCITA,references to it were conspicuously missing from the WorkingGroup report to the Commission.B. UCITA’s Implications for the CISGThe debates surrounding UCITA (and not necessarily theprovisions of UCITA as they stand in the official version) mightshed light on three issues confronting UNCITRAL: whether (orhow) to extend the scope of the CISG beyond pure goods; adaptingthe existing contract formation rules of the CISG (beyond merelyincorporating the electronic contracting provisions of the ModelLaw) to address newer contracting methods such as shrink-wrapor click-wrap used frequently in electronic commerce; andextension of the CISG into the consumer area. Each of these threeareas has been extremely controversial in the UCITA process.ScopeThe appropriate scope of UCITA has been a persistentproblem throughout its drafting history. As was noted earlier,work on information contracts in the United States began with anexamination of the scope of Article 2. Attempts to broaden Article2 to cover software were ultimately abandoned in favor of separatestatutory treatment of goods and information. As also notedearlier, the decision to separate goods and software was not onereached by the drafting committee, but by the Conferenceleadership, its decision representing a political or administrativedecision much more than a decision on the merits. After the split,the scope provisions of UCITA (and its predecessor, Article 2B)received a great deal of discussion throughout the remainingdrafting process, were the object of most of the criticism of thedraft by the American Law Institute, and have continued to benot adequate to address the way in which technology services and items suchas software were being sold.Id. The Secretariat’s report also noted two ICC projects: the draft Uniform Rules andGuidelines for Electronic Trade and Settlement (URETS) and the Model ElectronicSales Contract. Id. Elsewhere, the UNCITRAL staff member for the Working Groupon Electronic Commerce for UNCITRAL noted that UCITA “may serve as a usefulintroduction to electronic contracting issues that should be addressed, in addition tothose already included in the Model Law, on a global scale.” Renaud Sorieul et al.,Establishing a Legal Framework for Electronic Commerce: The Work of the UnitedNations Commission on International Trade Law (UNCITRAL), 35 Int’l Law. 107, 120(2001).

quite controversial.49 As the history of UCITA demonstrates,there was a two-fold recognition in the drafting process: that tosome extent the application of some of the sales provisions tosoftware and information products was appropriate, but that therewas a need for different rules in certain areas. The firstbattleground has been on where there are sufficient differences tojustify different rules i.e., which of Article 2’s rules could beapplied without change to software or information contracts.Industry has been adamant in its opposition to the expansion ofArticle 2 in any fashion to cover information products (includingsoftware).50 That opposition ultimately led to the split betweenArticles 2 and Article 2B. After the split, however, a secondbattleground developed that still persists: what transactions arecovered by the old sale of goods rubric (Article 2) and which aresubject to the new act (UCITA).After abandonment of attempts to include software productswithin the scope of the Article 2 of the Uniform Commercial Code,efforts were made to define the line between what constitutedgoods (covered by the UCC) and what constituted information(governed by Article 2B or UCITA).51 Since both Article 2 andUCITA contain scope provisions, two different but interrelateddebates flourished. Those supporting UCITA (and its substantiveprovisions) wanted as broad a scope provision as possible in thatact and as restrictive a scope provision as possible for Article 2; inaddition, they wanted to give the parties the ability to opt into thescope of UCITA through contract.52 Those opposing UCITA, of49. See supra notes 25, 27.50. See, e.g., Lorin Brennan, Why Article 2 Cannot Apply to SoftwareTransactions, 38 Duq. L. Rev. 459 (2000) (arguing, however, that either Article 2 mustapply in its totality or a separate statute must govern).51. Indeed, not only did the decision to split Article 2B off from Article 2 increasethe importance of the scope provisions, it turned the scope provisions of both drafts intoa “political football.” See Speidel, Symposium Intro, supra note 12, at 792 (“[A]fter thecollapse of “hub and spoke,” there was growing tension between the Article 2 and theArticle 2B projects (now UCITA), both as to the degree of textual conformity thatshould exist between them and the line beyond which a sale of goods stopped and acomputer information transaction began. That tension persists to this day.”); see alsoAnn Lousin, Proposed UCC 2-103 of the 2000 Version of the Revision of Article 2, 54SMU L. Rev. 913 (2001) (explaining the changes to the draft Article 2). For a critiqueof the current scope provisions of UCITA, see Stephen Y. Chow, UCITA: A 1990’sVision of E-Commerce, 18 J. Marshall J. Computer & Info. L. 323 (1999).52. The broad reach of UCITA, and the potential it might sweep in transactionsthat would otherwise be governed by Article 2, led to a motion at the 1998 ALI AnnualMeeting discussing Article 2B, the predecessor of UCITA, to “limit the scope of Article

course, wanted to keep its scope narrow and were opposed to anyattempt to limit the scope of Article 2 in any way. Some were ofthe view that software and information (particularly off-the-shelfsoftware) should not be excluded at all from the definition ofgoods, primarily on the grounds that this would deny purchasersprotections they previously had.53One issue was the mixed transaction: goods with an elementof information (either “embedded” in or packaged with the goods).Critics of UCITA expressed concern (i) that the presence ofsoftware or information would appropriately exclude somethingfrom Article 2 coverage; and (ii) that in appropriate circumstancesArticle 2 would cover both the information and the goods as aunit.54 By contrast, supporters of UCITA wanted to retain theability, to the extent possible, to have UCITA continue to cover theinformation and even the medium or tangible good on which theinformation was delivered.55 Extreme discomfort with the scope2B to software contracts, access contracts, and such other transactions as are includedexpressly and defined with sufficient clarity to avoid surprise to affected parties.” Themotion was withdrawn in light of representations by the Chair of the DraftingCommittee that a task force would be producing a report on the appropriate scope ofArticle 2B. Actions Taken with Respect to Drafts Submitted at 1998 Annual Meeting, athttp://www.ali.org/ali/ACTIONS.htm (last visited Jan. 13, 2002).53. The argument was that courts had “gotten it right” when the question hadarisen in the past, and thus it would be inadvisable to override existing law. See, e.g.,Stephen Y. Chow, Motions Relating to Proposed Revised Uniform Commercial CodeArticle 2 (1999), at http://www.ali-aba.org/ali/1999_CHOW_UCC2.htm (last visitedSept. 8, 2001); Memorandum from Jean Braucher & Peter Linzer to Members of theAmerican Law Institute (May 5, 1998), available at http://www.aliaba.org/ali/braucher.htm(last visited Sept. 8, 2001). At the 1998 ALI Annual Meeting,a motion was made to delete the scope provisions in the 1998 Article 2 draft that wouldhave excluded software from Article 2 coverage in order to remain silent on the issue;the motion was defeated by a vote of 97 to 78. See Actions Taken with Respect to DraftsSubmitted at 1999 Annual Meeting, available at http://www.ali.org (last visited Nov.17, 2001); see also Peter A. Alces, W(h)ither Warranty: The B(l)oom of ProductsLiability Theory in Cases of Deficient Software Design, 87 Cal. L. Rev. 271 (1999)(concluding that strict liability could replace the warranty protections in the UCC);Marion W. Benfield, Jr. & Peter A. Alces, Reinventing the Wheel, 35 Wm. & Mary L.Rev. 1405 (1994) (analyzing the hub and spoke approach with regard to contract lawand commercial law).54. The classic example is the automobile with a computer chip controlling theautomatic braking system. The consensus was that, in the event of a failure of thebraking system, the buyer of the car would be able to proceed under Article 2 and nothave her remedies depend upon proof of whether the hardware or the software was thecause. For a criticism of UCITA’s treatment of the mixed transaction, see Chow, supranote 51, at 331.55. UCITA as currently drafted covers not only computer software, but themedium on which it is stored, i.e., the “copy.” See UCITA § 103(b)(1).

provisions of Article 2 (based in part on the UCITA formulation)led to a straw vote of the membership of the Institute in 2000favoring a clear statutory statement that coverage of Article 2extends to so-called “smart goods,” products (such as a refrigeratoror a car) that include computer programs important to theirperformance.56 Efforts to produce an acceptable scope provisioncontinued to engender controversy, and, following an onlinediscussion forum hosted by the Institute, these efforts wereabandoned. The draft finally approved by the Institute in May2001 made no attempts at all to change the scope provisions toaddress computer software or information generally.57 Ironically,the failure to amend the scope of Article 2 has caused someinterest groups (particularly the software industry) to announcetheir opposition to the amendments to Article 2 as they go forwardto the National Conference for approval.58[This] Act treats the medium that carries the computer information as part ofthe computer information and within this Act, whether the medium is atangible object or electronic. This Act applies to the copy, documentation,and packaging of computer information; these are within the definition ofcomputer information itself. They are mere incidents of the transfer of theinformation.Id. at cmt. 4(b)(ii) (citation omitted).56. Article 2 Update, A.L.I. Rep. (Summer 2000), athttp://www.ali.org/ali/R2204_Update.htm (last visited Sept. 8, 2001); Actions Taken on2000 Annual Meeting Drafts, at http://www.ali.org (last visited Nov. 17, 2001). See alsoHeavy Agenda Planned for 2001 Annual Meeting, A.L.I. Rep. (Winter 2001), athttp://www.ali.org/ali/R2302_heavy.htm (last visited Sept. 8, 2001) (“The most difficultand controversial problem confronting the drafters of Revised Article 2 of the UCC hasbeen that of determining and defining the extent to which computer programsassociated with goods, particularly those contained in so-called ‘smart goods,’ should besubject to the provisions of Article 2.”).57. See Proposed Amendments for Uniform Commercial Code Article 2—Sales(May 2001), available at http://www.law.upenn.edu/bll/ulc/ucc2/ART20501.htm (lastvisited Sept. 8, 2001) (noting in the Prefatory Note to the draft: “These amendmentsalso reflect an inability to reach reasonable consensus on some issues, such as . . . theproper scope of the Article.”).58. See Letter from Microsoft to Commissioners of the National Conference ofCommissioners on Uniform State Laws (July 2001). The desire of industry to push fora carve-out, particularly for pure information, was undoubtedly spurred by cases suchas Specht v. Netscape Communications Corp., 150 F. Supp. 2d 585 (S.D.N.Y. 2001)(purporting to apply Article 2 to pure information contracts such as Internet downloadswhere only information and no goods are involved).At the August 2001 Annual Meeting of the National Conference, the softwareand information industries succeeded in convincing the Drafting Committee to changeits position, by a vote of 6-3, and recommend a new scope provision carving outinformation from the coverage of Article 2, and introducing new provisions to governmixed transactions. Later that day, the new provision survived a motion to delete on

Contract FormationThe second area of controversy that may have importantramifications on the CISG discussions on electronic contracting isthe treatment of contract formation issues in the context oflicensing, and in particular, its application to shrink-wrapsituations (where an item is purchased in a box or wrappingcontaining language that by opening or using the item, the userconsents to terms contained therein) as well as click wrapsituations (where a party signifies it acceptance of new oradditional terms by a mere mouse click). In many respects, thissecond set of issues deal with straight contract formation.Theoretically, contract formation rules should be the same, nomatter what the nature or subject matter of the transaction,59 yetthere was continual tension between the Article 2 and Article 2Bprocesses as to what those correct contract formation rules shouldbe. The contract formation rules of UCITA have proven to bequite controversial, in large part because of their impact uponintellectual property law regimes.60 Even apart from theircontent, however, the approach exemplified by UCITA’s provisionshas been challenged, and indeed they have not been mirrored inthe Article 2 contract formation provisions.61UCITA legitimates both shrink-wrap and click-wrapagreements.62 Key to its provisions is the concept of “manifestingassent” after an “opportunity to review” the terms of a standardthe floor of the Conference by a vote of 60 to 98, but a motion necessary to obtainNCCUSL approval of the entire project (including the new scope provision) failed on avote of 53 to 89. This leaves the issue of the scope of revised Article 2, and indeed thelarger question of whether there will ever be a revised or amended Article 2,unresolved.59. As Judge Easterbrook said in Hill v. Gateway 2000, Inc., 105 F.3d 1147, 1149(7th Cir. 1997), cert. denied, 118 S. Ct. 47 (1997) (involving a purchase of a computerwhich was delivered with new and additional terms), about attempts to distinguish theprior case of ProCD v. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996) (involving purchasedsoftware which came with a shrink-wrap license limiting use): “ProCD is about the lawof contract, not about the law of software.”60. See infra notes 123-44 and accompanying text; see also Brennan, supra note50.61. See, e.g., Speidel, Trenches, supra note 12; Rusch, supra note 12.62. Most shrink-wrap or click-wrap agreements would qualify as “mass market”licenses under UCITA’s provisions, and their enforceability would be governed byUCITA § 209. See UCITA § 112 & official cmt. 5 (setting forth several illustrations ofthe enforceability of click-wrap agreements).

form license of information.63 Although there is generalagreement that parties should be able to contractually agree tothe terms applicable to their transaction,64 the continuing refrainheard from opponents is that under UCITA there is no truebargain and no true “meeting of minds,” and thus there is nocontract in a classical sense.65 Such fictional assent is arguablypresent in a host of circumstances leading to a binding agreementunder UCITA. For example, assent may be found by a simple“click of the mouse”—or a double-click—when a person isattempting to access information. Thus, the person may not haveread the terms, known they were present, or even known that the“click” was a binding agreement, and still be bound to thetransaction.66 Assent may be found where an individual clicksthrough one screen on a website, even though the governing termsare not themselves displayed on the website but are simplyavailable should the individual choose to click through to anotherpage. The terms to which the party agreed may allow the licensor63. UCITA provides that an opportunity to review is presented if licensor displays“prominently and in close proximity to a description of the computer information, or toinstructions or steps for acquiring it, the standard terms or a reference to an electroniclocation from which they can be readily obtained.” UCITA § 211. Alternatively, thelicense can “disclos[e] the availability of the standard terms . . . on the site . . . andpromptly furnish[] a copy of th[os]e standard terms on request before the transfer ofthe computer information . . . .” Id. In either case, the licensor is not permitted toprevent the printing or storage of the standard terms for archival or review purposesby the licensee. Id.64. This position was advanced by the Clinton administration as critical to thedeveloping framework for electronic commerce. Clinton & Gore, Framework, supranote 7 (“In general, parties should be able to do business with each other on theInternet under whatever terms and conditions they agree upon.”). Freedom of contracthas been one of the key foundations of UCITA. See Carlyle C. Ring, Jr., Uniform Rulesfor Internet Information Transactions: An Overview of Proposed UCITA, 38 Duq. L.Rev. 319 (2000).65. See, e.g., Jean Braucher, Uniform Computer Information Transactions Act(UCITA): Objections From The Consumer Perspective, 5 Cyberspace Law., Sept. 2000,at 2, available at http://www.cpsr.org/program/UCITA/braucher.html (UCITA“validates fictional assent (e.g., double clicking a mouse to get access to a product afteryou’ve paid for it) and even allows one party to define any conduct as assent in futuretransactions, without requiring that form terms meet consumers’ reasonableexpectations.”); Chow, supra note 51, at 337.66. Under UCITA, a person “manifests assent” if the person, after having anopportunity to review the record or term, intentionally engages in conduct with reasonto know that the other party assents from that conduct. UCITA § 112(a). It is notnecessary that the person know whether the record or term exists; rather he or shesimply needs an opportunity to review. Nor is it necessary for the person to know thattheir conduct constitutes assent; there need only be a “reason to know.”

to define any conduct as assent in the future, or it may allow thelicensor in access contracts (involving electronic access toinformation on another’s system)67 to unilaterally change theterms of the agreement. Thus, the concept of “manifestation ofassent” has been attacked as a mere fiction.68 Indeed, recentevents have demonstrated that not only are licensees frequentlyunaware of the terms in click-wrap agreements, but sometimeseven the licensors are unaware of what they have in their ownstandard form provisions, with the result that “implied assent” iscarried to a new level.69Whatever one thinks of the “assent agreement” underdomestic United States law, there is the additional problem of itscompatibility with legal principles internationally. The expansiveconcept of personal autonomy exemplified in UCITA has also beencriticized as contrary to non-U.S. concepts of assent.70Additionally, the UCITA provisions seems to conflict with67. See UCITA § 102 (a)(1).68. But see Raymond T. Nimmer, International Information Transactions: AnEssay on Law in an Information Society, 26 Brook. J. Int’l L. 5, 44-45 (2000).Some have argued that a double click on an ‘I agree’ icon is only ‘fictional’assent that should be ignored. . . . Most likely, however, this argument isdisingenuous; the persons using it actually believe that parties should not bebound by standard forms, even in the online world. But in Internetcommerce, all contracts involve standard forms. Does that then indicate thatcontractual terms cannot be agreed to online? The consequences of such arule would be draconian and would contradict the basic notion of contract andthe economic reality of modern information commerce.Id.69. A recent incident involved Microsoft’s license for Passport, which purported toclaim an ownership interest in all personal data passing through the system, thusgranting Microsoft the right to “use, modify, copy, distribute . . . or sell” all personaldata. When this draconian term (which had been included in the license for some time)finally came to light, Microsoft responded that it was unaware of the term in its ownlicensing agreement. See Farhad Manjoo, Fine Print Not Necessarily in Ink, WiredNews (Apr. 6, 2001), available athttp://www.wired.com/news/print/0,1294,42858,00.html (last visited Sept. 8, 2001).The irony is, of course, that without judicial intervention setting aside the terms in theMicrosoft license, the licensor and the licensee have agreed to terms of which neitherare aware! Yet there has been the creation of an agreement under UCITA.70. This is especially true in the intellectual property area. See Samuel K.Murumba, The Emerging Law of the Digital Domain and the Contract/IP Interface: AnAntipodean Perspective, 26 Brook. J. Int’l L. 91, 113 (2000) (“UCITA itself rests less onany utilitarian conception and more on that most neo-Kantian and neo-classical of allconceptions: autonomy. There is an unmistakable resistance in Australia, and it seemsin Europe, to a wholesale embrace of this super-naturalistic view of intellectualproperty rights, whether it comes in the guise of autonomy or code or both.”).

principles articulated in several international instruments. Forexample, a directive of the European Union deems aspresumptively unfair a term “irrevocably binding the consumer toterms with which he had no real opportunity of becomingacquainted before the conclusion of the contract.”71 Also deemedunfair are terms enabling the supplier to alter the terms of thecontract unilaterally without a valid reason specified in thecontract.72In addition, UCITA for the first expressly validates postpaymentdisclosure of terms—“pay now, terms later”73—underwhat is known as the “rolling contract” theory of contractformation. While it may not always be possible to disclose (ormake available) all applicable terms prior to the creation of abinding agreement (as in the case of a telephone purchase order),that is not the case with respect to online transactions whereinformation can be made readily accessible. Indeed, one can arguethat the Internet is the quintessential marketplace: a place wherethe notional or theoretical fully-informed party to a transactioncan be more closely approximated than in physical markets; anenvironment capable of providing the information that buyersneed to make informed shopping decisions.74 Rather thanrecognizing the ability of the media to resolve informationaldistribution issues, however, UCITA condones the delay ofdisclosure of terms until a consumer is committed to the deal, nomatter how important the terms.75 This is contrary to theapproach in the European Union, where prior disclosure of terms71. Council Directive 93/13/EEC of 5 Apr. 1993 on Unfair Terms of ConsumerContracts, 1993 O.J. (L 095) 29, Annex 1(i).72. Id. at Annex 1(j).73. Maureen A. O’Rourke, Progressing Towards a Uniform Commercial Code forElectronic Commerce or Racing Towards Nonuniformity?, 14 Berkeley Tech. L.J. 635,651 (1999).74. See id. at 652 (“Would it not be consistent . . . to draft legal rules that enhancethe market’s efficiency by providing more information to customers, correcting forinformation asymmetries that might otherwise exist and distort marketperformance?”); see also Jean Braucher, Delayed Disclosure in Consumer E-Commerceas an Unfair and Deceptive Practice, 46 Wayne L. Rev. 1805 (2000) (arguing thatmarketing to consumers online and failing to make pre-transaction disclosuresamounts, like other bait and switch practices, to unfair and deceptive practices inviolation of the Federal Trade Commission Act and states’ little FTC acts).75. See Braucher, supra note 65. UCITA does, however, contain a “safe harbor” toencourage posting. UCITA § 211.

is mandated in certain instances.76 As stated in Guidelines of theOrganisation for Economic Co-Operation and Development:“Businesses engaged in electronic commerce should providesufficient information about the terms, conditions and costsassociated with a transaction to enable consumers to make aninformed decision about whether to enter into the transaction.”77The easier it is to find or imply assent to the terms of acontract, the greater the pressure becomes to impose a method toprevent unfair or overreaching terms in the resultingagreements.78 The question of whether to enforce shrink-wrap orclick-wrap licenses is related to a second problem: whattechniques are there for policing unfair or overreaching termsincluded in the license by the licensor?79 To the extent that theseclick-wrap agreements are adhesion contracts, the user or licenseeeffectively has no choice but to accept the objectionable term orreject the entire transaction. To the extent the user has alreadypurchased a product and then finds additional terms (either undera rolling contract theory or because the licensor changes the termsafter the conclusion of the contract), the licensee may find itselfbound by terms it had no ability to negotiate.Repeated attempts in the Article 2 revision process to dealwith unfair terms in adhesion contracts were constantly objected76. Council Directive 97/7/EC of 20 May 1997 on the Protection of Consumers inRespect of Distance Contracts, 1997 O.J. (L 144) 19. A basic purpose of the directive isconsumer protection through prior information, written confirmations, right ofwithdrawal, performance, payment by card, and prevention of inertia selling.77. Organisation for Economic Co-Operation and Development, ConsumerProtection in the Context of Electronic Commerce Guidelines, § III(C) (Dec. 9, 1999)[hereinafter OECD], available athttp://www.oecd.org/oecd/pages/home/displaygeneral/0,3380,EN-document-0-nodirectorate-no-24-320-0,FF.html.78. The concept of contractual free choice and assent often sound good to thebusinessperson:However, contract law has a darker side as well. Cyberspace is not anegalitarian society with equal chances for every ‘netizen.’ In a world totallyruled by contract, weaker parties risk being subjugated and fundamentalfreedoms may be jeopardized. Freedom of contract may become contractualcoercion, especially when dominant undertakings abuse their market powerto impose contractual rules on powerless consumers, as if they were publicauthorities.P. Bernt Hugenholtz, Copyright, Contract and the Code: What Will Remain of thePublic Domain, 26 Brook. J. Int’l L. 77, 79 (2000).79. See John J.A. Burke, Contract as Commodity: A Nonfiction Approach, 24Seton Hall Legis. J. 285 (2000) (surveying the policing approaches available tostandard form contracts and recommending a regulatory response).

to by industry groups, with the result that ultimately no changeswere included in the 2001 final amendments to Article 2.80UCITA deals with the objectionable term in the license in twoways.81 First, any term that is found to be “unconscionable” willbe unenforceable.82 Second, if the term is available to the licenseeonly after a person becomes obligated to pay or beginsperformance, the licensee after having an opportunity to reviewthe license may if its chooses return the information and obtain arefund.83 Although allowing an opportunity to return has been80. See, e.g., section 2-206(b) of the November 1, 1996, Council Draft (providingthat where a consumer manifests assent to a “standard form, a term contained in theform which the consumer could not have reasonably expected is not part of contractunless the consumer expressly agrees to it.”). In the July 1999 draft, subsection (a) tosection 2-206 provided that “in a consumer contract, a court may refuse to enforce astandard term in a record the inclusion of which was materially inconsistent withreasonable commercial standards of fair dealing in contracts of that type, or . . .conflicts with one or more nonstandard terms to which the parties have agreed.”Commercial interests opposed all versions of new section 2-206, argued that theproblem was already adequately addressed in the unconscionability provisions of UCCsection 2-302, and insisted that any efforts to particularize UCC section 2-302 shouldbe in the comments. See Speidel, Trenches, supra note 12 (calling attempts to draft anew provision particularizing the elements of unconscionability for consumer contractsthe “pea under the mattress”). See also infra note 82.81. In the drafting of UCITA (as well as the drafting of amendments to Article 2)attempts were made to deal with adhesion contracts generally. Ultimately, thoseattempts were abandoned, primarily because of opposition from licensors. The onlypolicing mechanism is that of unconscionability. The second method, discussed below,deals only with terms that come after the obligation to pay.82. UCITA § 111. There has been a continuing argument about whether theunconscionability concept found in UCC section 2-302 gives sufficient protectionagainst onerous terms in adhesion contracts. As a result, the 1999 Article 2 AnnualMeeting Draft contained the following addition to section 2-302, which was approvedby a 2-1 margin:(b) In a consumer contract, a nonnegotiated term in a standard form record isunconscionable and is not enforceable if it(1) eliminates the essential purpose of the contract;(2) subject to Section 2-202, conflicts with other material terms to which theparties have expressly agreed; or(3) imposes manifestly unreasonable risk or cost on the consumer in thecircumstances. Id.When the 1999 Annual Meeting Draft was withdrawn from consideration, the provisionwas abandoned and not carried over by the new Drafting Committee into what are nowthe 2001 Amendments to the Article 2, in large part because of the organizedopposition of industry. See Speidel, Trenches, supra note 12. Thus, the question ofpolicing standard terms in adhesion contracts has been left to the courts on a case-bycasebasis.83. UCITA § 112 (explaining that a person can manifest assent to a term or recordonly if he or she has had an opportunity to review its terms, which in the case of termscoming after the obligation to pay arises, means only if the person has a right to a

lauded by some as a step forward for consumer protection, theobjectors note that the right of return is of questionable value: theright of return can be easily obviated by the licensor who makesthe information available ahead of time by posting it on thewebsite,84 thus ignoring the fact that the terms are non-negotiableand unread; the right does not apply if the new terms areconstrued as proposed modification of terms;85 and most users orlicensees never read the terms in the license anyway. Even if theyread the term, the term at the time of acquisition may not seemsufficiently important to merit a return; the costs invested in dealto date may be great; and transaction costs (time, aggravation,replacement) are not covered.Again, these types of provisions must be examined againstinternational concepts of appropriate policing tools, where theconcept of policing against unfairness is seen as a business issueas much as a consumer issue.86 The most relevant provision laidout in the UNIDROIT Principles of International CommercialContracts allows a court to set aside a term in a standard form ifit is “of such a character that the other party could not reasonablyhave expected it . . . .”87 Ironically, this formulation, which theUNIDROIT Principles apply to commercial contracts, was rejectedby the UCITA drafting committee as too liberal, even in theconsumer context. Other international documents give additionalpowers to courts to police transactions in the consumer context,return if he or she rejects the term). This concept is further elaborated in the case ofmass market licenses in UCITA section 209.84. UCITA §§ 112, 212 (if there is pre-transaction disclosures of contract terms inInternet transactions where the contract is formed online for an electronic delivery ofinformation then an opportunity to review is deemed given).85. Id. § 112.86. See Michael J. Bonell, An International Restatement of Contract Law: TheUNIDROIT Principles of International Commercial Contracts 90 (Rome, 1994).An optimistic assumption that long prevailed at both domestic andinternational level was that contracts between merchants are concluded onlyby experienced and competent professionals in acting in accordance withwell-established principles of fair dealing . . . . This assumption isincreasingly being questioned . . . the UNIDROIT Principles move from amore realistic evaluation of international commercial bargains and provide avariety of means for ‘policing’ the contract or its terms against bothprocedural and substantive unfairness.Id.87. UNIDROIT Principles of International Commercial Contracts (1994), art.2.20(1), available at http://www.unidroit.org/english/principles/pr-main.htm.

powers going well beyond the UNIDROIT rules.88The combination of the assent rules of UCITA (which find abinding agreement at the slightest click of a mouse) and thepolicing rules which do not give adequate oversight againstoverreaching led to passage of the following sense-of-the-housemotion at the 1998 ALI Annual Meeting discussion on Article 2B(UCITA’s precursor): “The current draft of proposed UCC Article2B has not reached an acceptable balance in its provisionsconcerning assent to standard form records [§§2B-111, 2B-203,2B-207, 2B-208, and 2B-304] and should be returned to theDrafting Committee for fundamental revision of the severalrelated sections governing assent.”89 Of course, membership ofthe ALI never evaluated the final act (renamed UCITA) againstthis motion to police compliance with its wishes as a result of thedisengagement of the ALI from the process in the spring of 1999.Consumer TransactionsThe third area of controversy with potential relevance to theCISG debate involves coverage of consumer transactions withinthe CISG. The need to cover consumer as well as commercialtransactions under one international legal rubric stems from animportant feature of electronic commerce: on the Internet, theretail and wholesale markets converge, and the ability todistinguish between consumers and merchants diminishes. Werethe CISG to cover consumer transactions, there wouldundoubtedly be both the risk that the CISG would undermineexisting consumer protection under national laws, and the riskthat pressure would be brought in the revision process to dealwith consumer protection in the convention itself (a very difficulttask given the disparate nature of consumer protection regimesaround the world). In the drafting of UCITA, consumer protectionwas a significant area of concern,90 sparking opposition by, among88. See infra notes 93-104.89. Braucher & Linzer, supra note 26.90. See, e.g., Ajay Ayyappan, UCITA: Uniformity at the Price of Fairness?, 69Fordham L. Rev. 2471 (2001) (discussing the warranty provisions of UCITA and theirimpact on fairness); Christopher T. Poggi, Electronic Commerce Legislation: AnAnalysis of European and American Approaches to Contract Formation, 41 Va. J. Int’lL. 224 (2000) (urging efforts to develop international conventions on electroniccommerce, particularly in the consumer area, but noting that U.S. proposals such asUCITA are “far from uncontroversial”); Michael L. Rustad, Making UCITA More

others, the Federal Trade Commission and the NationalAssociation of Attorneys General.91 For example, the failure ofUCITA to deal with such matters as prior online disclosure ofterms has been criticized as a failure to recognize applicablenorms in consumer protection.92Traditional contract law, in the United States and otherjurisdictions, has developed doctrines that allow courts to setaside contracts, in whole or in part, if their terms (or one of theirterms) are objectionable. Outside the United States, the lawappears to be more protective against the application of unfaircontract terms. The UNIDROIT Principles of InternationalCommercial Contracts is a prime example, barring theenforceability of terms in standard form contracts that are bothunreasonable and “surprising.”93 It is noteworthy that theUNIDROIT Principles are not consumer protection principles; bytheir own terms, the Principles apply only to commercialcontracts.94 Yet, this formulation for governing terms in standardform contract was rejected during the drafting of UCITA (and therevisions to Article 2 as well).95The European Union has enacted two regulations that giveadded protection to consumers, protection not similarly afforded inUCITA. The EU Directive on Unfair Terms in ConsumerConsumer-Friendly, 18 J. Marshall J. Computer & Info. L. 547, 581 (1999) (“If UCITAis to serve as an international template for licensing software, [it] should at leastincorporate some additional consumer protections.”); Saami Zain, Regulation of E-Commerce by Contract: Is It Fair to Consumers, 31 UWLA L. Rev. 163 (2000)(examining deficiencies in contractual approach, and in UCITA, with regard toconsumer concerns in e-commerce).91. See Letter from State Attorneys General to Gene N. Labrun, supra note 18.92. O’Rourke, supra note 73, at 653.Over the years, federal and non-uniform state enactments in the consumerprotection area effectively either preempted or modified parts of Article 2,reflecting emerging views of relevant differences between business-tobusinesstransactions and business-to-consumer transactions. Article 2B hasthe opportunity to incorporate consumer protection into its text, avoiding thelater preemption that characterized Article 2.Id.93. See supra note 87.94. UNIDROIT Principles of International Commercial Contracts (1994),Preamble, available at http://www.unidroit.org/english/principles/princ.htm#NR1 (lastvisited Nov. 17, 2001) (“These Principles set forth general rules for internationalcommercial contracts. They shall be applied when the parties have agreed that theircontract be governed by them.”).95. See, e.g., Speidel, Trenches, supra note 12; Rusch, supra note 12.

Contracts deals with unfair terms in sales contracts with aconsumer buyer. Under that Directive, a term is unfair if (i) it hasnot been individually negotiated;96 and (ii) it causes a significantimbalance in the parties’ rights and obligations arising under thecontract, to the detriment of the consumer.97 The annex to thedirective lists some terms that are prima facie unfair, includingsome terms ostensibly condoned by UCITA.98 Terms must be inplain and intelligible language.99 Furthermore, where there isdoubt about the meaning of a term, the interpretation mostfavorable to the consumer shall prevail.100 The EuropeanDirective on Distance Selling Contracts grants certain substantivesafeguards in “distance contracts.”101 These safeguards includerequiring certain disclosures prior to formation, in addition togranting consumers a seven-day unilateral right to withdraw.102Choice of law clauses which deny consumers the benefit of such96. Council Directive 93/13/EEC of 5 Apr. 1993 on Unfair Terms of ConsumerContracts, 1993 O.J. (L 95) 29, art. 3.2 (“A term shall always be regarded as notindividually negotiated where it has been drafted in advance and the consumer hastherefore not been able to influence the substance of the term, particularly in thecontext of pre-formulated standard contract.”).97. Id. at art. 4 (“[T]he unfairness of a contractual term shall be assessed, takinginto account the nature of the goods or services for which the contract was concludedand by referring, at the time of conclusion of the contract, to all the circumstancesattending the conclusion of the contract . . . .”).98. Examples of possibly unfair terms under the Directive potentially conflictingwith UCITA include: inappropriate exclusions or limitations on a consumer’s remediesagainst the seller or supplier, including a right of offset granted the seller (Annex 1(b));terms requiring a breaching consumer to pay a disproportionately high sum incompensation (Annex 1(e)); terms allowing the seller to terminate a contract ofindefinite duration without reasonable notice except where there are serious groundsfor doing so (Annex 1(g)); terms irrevocably binding the consumer to terms with whichhe had no real opportunity of becoming acquainted before the conclusion of the contract(Annex 1(i)); and terms allowing the supplier to unilaterally change the terms of thecontract without a valid reason specified in the contract (Annex 1(j)).99. Council Directive 93/13/EEC of 5 Apr. 1993 on Unfair Terms of ConsumerContracts, supra note 96, at art. 5 (stating “where all or certain terms offered to theconsumer are in writing, these terms must always be drafted in plain, intelligiblelanguage”).100. Id.101. Council Directive 97/7/EC of 20 May 1997 on the Protection of Consumers inRespect of Distance Contracts, 1997 O.J. (L 144) 19. A distance contract means anygoods or services contract “concluded between a supplier and a consumer under anorganized distance sale . . . scheme run by the supplier, who, for the purpose of thecontract, makes exclusive use of one or more means of distance communication” priorto contract creation. Id. at art. 2(1). This would include contracts created by e-mail aswell as those concluded electronically.102. Id. at arts. 4-6.

ights are suspect.103 Other attempts to protect consumersagainst overreaching in adhesion contract situations have beenproposed in Europe.104 Any attempts to enlarge the scope of theCISG, or to draft another international instrument dealing withboth commercial and consumer transactions, will need to come toterms with these norms, which are broadly accepted in majortrading nations.C. UCITA as an International Uniform ActAn alternative to enlarging the scope of the CISG toaccommodate software or information contracts is to follow theUCITA precedent and develop an international uniform lawdealing with such transactions. Again, if such an effort wereundertaken, there is much to learn from UCITA and thediscussions surrounding its drafting and adoption. It is highlyunlikely, however, that UCITA would ever be found amenable towholesale adoption on the international level.105Structural IssuesThere are a number of observations (both general andspecific) that support this conclusion. First, some say a picture is103. Id. at art. 12.Member States shall take the measures needed to ensure that the consumerdoes not lose the protection granted by this Directive by virtue of the choice ofthe law of a non-member country as the law applicable to the contract if thelatter has close connection with the territory of one or more Member States.Id.104. In a recent letter to Parliament, the Dutch Minister of Justice suggestedintroducing so called “unwaivable use rights” in order to expressly protect informationconsumers against unconscionable licensing practices. Hugenholtz, supra note 78, at83 n.27; see also Francois Dessemontet, The European Approach to E-Commerce andLicensing, 26 Brook. J. Int’l L. 59, 68 (2000) (“[T]he French Conseil d’Etat hassuggested that a hyperlink be created so that the professional codes of conduct beembodied as General Terms and Conditions of the contract . . . . Such a renewedconfidence in General Terms and Conditions could allow for the European Law and theU.S. self-regulatory approach to converge to a larger extent.”) Arguments made duringthe UCITA drafting process, however, that certain terms (e.g., choice of law rules)should be “non-waivable” in mass market licenses were rejected.105. A forthcoming German book examines the appropriateness (orinappropriateness) of the UCITA model under German law, considering such areas aswarranties, choice of law, and formation. Der E-Commerce Vertrag nach US RechteineEinfhrung in den Uniform Computer Information Transaction Act (UCITA) unterBercksichtigung der Auswirkungen auf das Vertragsrecht fr IT Produkte (OttoSchmidt Verlag, forthcoming 2002).

worth a thousand words. Picture a draft 261 single-spaced pageslong (or 123,829 words), consisting of eight parts with a total ofone-hundred four (104) substantive provisions.106 Its shear lengthand complexity makes UCITA a difficult act to understand, evenfor those who are familiar with its provisions. Additionally, thetechnical nature of much of the material and the detaileddefinitions (66 in total—many with substantive rules in thedefinitions) combine to make reading and understanding the act achallenge. Compounding the problem is the drafting style that isoften impenetrable, even for those familiar with UCITA. Anexample is section 601(d): “Except as otherwise provided in section603 and 604, in the case of a performance with respect to a copy,this section is subject to sections 606 through 610 and sections 704through 707.”107There are two factors that contributed to the long,comprehensive nature of the UCITA draft, which has ambitiouslytried to cover virtually every issue raised during the draftingprocess. Although the original project set out only to coversoftware, and UCITA, as finalized, covers only computerinformation transactions, drafts during the intervening stageswere much broader in scope, covering all transactions ininformation.108 This broad scope brought a wide variety ofindustries to the drafting table, all with their own agendas andissues; many provisions were incorporated into the final draft inan attempt to build support for the Act.109 When the scope wasnarrowed, many of these provisions were left in.110 Even with its106. If one considers the transition rules as well, there are 109 provisions in nineparts.107. UCITA § 601(d).108. See Pamela Samuelson & Kurt Opsahl, How Tensions Between IntellectualProperty Policy and UCITA Are Likely to be Resolved, in Patents, Copyrights,Trademarks, and Literary Property Course Handbook Series 741, 747 (1999) (“Initiallyonly software was involved, but then on-line databases came onboard, followed bydigital information products and services such as CD-ROMs. By 1995, the scope of theproposed law . . . had expanded to cover information licensing generally.”).109. There was a second tendency that arose when an industry raised substantialobjections to the draft: eliminate them from the act’s coverage. See UCITA § 103(d)(excluding, inter alia, financial services transactions; insurance services transactions;motion picture or audio or visual programming; sound recordings, musical works, orphonorecords; and telecommunications products or services).110. UCITA has been analogized to a huge Christmas tree with ornaments forevery possible industry; the problem is that when the scope was changed and theChristmas tree was cut back, the number of ornaments remained the same.

narrowed scope, however, there are those who have pointed outthat the multiplicity of types of transactions covered by UCITA(ranging from off-the-shelf purchases of software, to accesscontracts, to contracts and to design software or databases) makesa “one size fits all” strategy unworkable.111 “The net result is that(UCITA), in trying to be all things to all types of computertransactions, has reached a level of complexity that is at oddswith . . . [the] goals of simplicity and clarification.”112A second factor contributing to the length and complexity wasthe desire to include provisions that were deemed essential to“sell” UCITA to legislatures and others. One such set ofprovisions are those dealing with contracting electronically (e.g.,using electronic mail or the Internet).113 Ironically, at the sametime UCITA was being drafted, the National Conference wasworking on a separate product—the Uniform ElectronicTransactions Act114—covering precisely those issues: electroniccontracting. Questions were repeatedly raised during the draftingprocess about the need to include such provisions in UCITA inlight of this separate Conference product;115 the justificationarticulated for keeping these provisions in UCITA was that itwould be better for those under UCITA to have all contractingrules included in one place. A more likely explanation for thereticence to remove these provisions is that these were exactly the111. See O’Rourke, supra note 73, at 648 (noting to the extent that the productconsists of a bundle of rights granted under a license, the contract and not anyobjective tangible characteristics defines the product; the result is that a change incontract terms changes the product and results in a multiplicity of types of contracts).112. Id. at 647.113. By contrast, those developing other types of e-commerce legislation found itimportant to distinguish between the rules governing electronic contracting in general,and the substantive rules governing the underlying transaction. Thus, the UNCITALModel Law on Electronic Commerce and the Uniform Electronic Transactions Act focussolely on the use of an electronic medium to contract and do not attempt to combinethose issues with the substantive ones. See also O’Rourke, supra note 73, at 648(noting that the unifying factor in the European E-Commerce Directive was the meansof contracting rather than the subject matter of the contract, and suggesting thatUCITA might “benefit from some synthesis” with the Commission Directive).114. The Uniform Electronic Transactions Act (UETA) was adopted by theNational Conference of Commissioners on Uniform State Laws in August of 1999. Thefull text of the act, with comments, is available athttp://www.law.upenn.edu/bll/ulc/fnact99/1990s/ ueta99.htm (last visited Nov. 17,2001). As of September 2001, the UETA had been enacted in 38 states. Up-to-dateinformation on the enactment status of the UETA may be found athttp://www.nccusl.org/uniformact_factsheet/uniformacts-fs-ueta.asp.115. Indeed, the ALI was one of those questioning the need for those provisions.

types of provisions being sought by many major industries seekingcertainty regarding the enforceability of electronic contracts. Byincluding these needed provisions, the hope was that those doingelectronic contracting would support the draft as a whole.116 Theimportance of these electronic contracting provisions is illustratedby the argument repeatedly made during the drafting process thatUCITA needed to be finished quickly to avoid federalpreemption.117 The only major threat of preemption from inactionwas in the field of electronic contracting, and that was realizedwith the passage of the federal E-Sign legislation.118The exceptional detail in UCITA is a substantial barrier to itsutility on the international level. Even if such detail might bewarranted in the United States, for domestic and legislativepurposes, it would be difficult to reach consensus on suchtechnical matters internationally, given the divergent legalsystems that need to be accommodated. Moreover, many of theprovisions of UCITA were written against the backdrop of U.S.law,119 and the resulting rules may not be consistent with otherlegal systems.In addition, UCITA was built against a backdrop of existingtrade usages and practices—usages and practices that may not berepresentative of those occurring on the international level, orindicative of those that may exist in the future. Indeed, one of the116. There are those who have cynically observed that most people are unaware ofthe distinctions between UCITA and UETA, frequently confusing the two, and thatthose who have indicated support for the former may well have only intended tosupport the latter! There are undoubtedly substantive reasons why the UCITADrafting Committee wanted to retain its electronic contracting provisions. Bycomparison to the UETA drafting committee, the UCITA Drafting Committee was infavor of detailed rules that would strengthen the ability of parties (e.g., licensors) toenforce electronic contracts. See Amelia H. Boss, Searching for Security in the Law ofElectronic Commerce, 23 Nova L. Rev. 585 (1999). Thus, UCITA’s proposed rules onattribution were more detailed and directive than the corresponding provisions in theUETA. Ultimately, however, UETA’s formulation prevailed and was adopted forUCITA on the floor of the National Conference.117. This argument was made frequently by the Article 2B Reporter and DraftingCommittee Chair during presentations to both the ALI and the National Conference.118. Electronic Signatures in Global and National Commerce Act of 2000, 15U.S.C.A. §§ 7001-07 (2001). Ironically, the passage of UCITA has spurred somepressure to respond with federal legislation or regulation preempting UCITAprovisions, at least in the consumer area. See, e.g., supra note 22.119. For example, the provisions on informational content (which were not uncontroversial)were drafted against the backdrop of the United States Constitution andits protection of free speech.

persistent questions raised about UCITA is whether, even withinthe United States, industry practice is sufficiently established tojustify codification.120 One of the reasons given for the decision toabandon attempts to include information within the UniformCommercial Code (Article 2B) and to proceed with a uniform lawoutside the Code was that the industry was sufficiently in fluxthat it would be premature to include it within the Code. Thissame reasoning may well justify not proceeding with it at all,domestically or internationally, at this time.121It should be noted that, on an international level, not allconcepts included in UCITA may be necessary or desirable. To theextent UCITA purports to cover electronic contracting, forexample, much of it may be unnecessary in light of products suchas the UNCITRAL Model Law on Electronic Commerce or theUNCITRAL Model Law on Electronic Signature. Moreimportantly, as noted earlier, many of the concepts of UCITA maybe inconsistent with international standards. Much of theconsumer treatment, for example, must be considered in light ofinternational norms such as those evolving within the OECD.122Collision With Intellectual PropertyProbably the greatest obstacle to the wholesale adoption ofUCITA internationally is its intrusion into the area of intellectualproperty. Although UCITA purports to be simply a “contracting”statute rather than an “intellectual property” regime,123 one of thebig areas of controversy in the United States is the intersection ofUCITA and its validation of licensing on intellectual propertyregimes.124 The relationship of licensing to intellectual property120. See O’Rourke, supra note 73, at 651 (“In many industries which would begoverned by [UCITA], it is premature to refer to a usage of trade. Customs are rapidlyevolving, and deference to a particular norm at a particular time may not beappropriate.”). Additionally, the technologies themselves are still rapidly changing.See A. Michael Froomkin, Article 2B as Legal Software for Electronic Contracting-Operating System or Trojan Horse?, 13 Berkeley Tech. L.J. 1023, 1026 (1998) (“Onereason why Article 2B [and UCITA have] proven to be so difficult to get right is thatthe information technologies to which [they] would apply are themselves in a state offerment.”).121. Commercial law tends to be descriptive or enabling rather than normative;until there is something to describe, it is difficult to legislate.122. See OECD, supra note 77 and accompanying text.123. See, e.g., Nimmer, Looking Glass, supra note 9, at 304 (“UCITA deals withcontracts and not property rights.”).124. Hugenholtz, supra note 78, at 78-79; Rochelle C. Dreyfuss, UCITA in the

law is a complex one.125 Although some have argued that a licenseis merely an agreement of the licensor not to sue for any violationof its intellectual property rights (and therefore merely supportsthe existing intellectual property regime), modern licensingschemes go much further: licensors use licenses to get more thanthey are granted by intellectual property law by increasing thesanctions against licensees for violation of that law; by placinglimitations on the rights granted to licensees by applicableintellectual property law (e.g., on users’ rights in or use of theinformation or by avoiding the first sale doctrine);126 and bycreating limitations on use of information that is not evenprotected by intellectual property law (e.g., non-copyrightableinformation).127 In addition, to get even more protection than isgiven under applicable intellectual property law and contract law,licensors have sought the technological ability to restrict the use ofinformation (both copyrighted and non-copyrighted), legislativeauthorization of these technological means,128 and legislativepenalization of those who attempt to circumvent thosetechnological means.129 These efforts have been criticized both onthe basis of conflicts with intellectual property policy andInternational Marketplace: Are We About to Export Bad Innovation Policy?, 26 Brook. J.Int’l L. 49, 52-53 (2000) [hereinafter Dreyfuss, Marketplace]; Rochelle C. Dreyfuss, DoYou Want to Know a Trade Secret? Licensing Under Article 2B of the UniformCommercial Code, 87 Cal. L. Rev. 1193 (1999) [hereinafter Dreyfuss, Trade Secret].125. A full exploration of that relationship is beyond the scope of this article. For aglimpse into the problem, see Rice, supra note 11.126. See David A. Rice, Licensing the Use of Computer Program Copies and theCopyright Act First Sale Doctrine, 30 Jurimetrics J. 157 (1990); Mark A. Lemley,Intellectual Property and Shrinkwrap Licenses, 68 S. Cal. L. Rev. 1239 (1999).127. See generally ProCD, Inc. v. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996)(involving a license in a non-copyrightable compilation of telephone numbers).128. See, e.g., World Intellectual Property Organization Copyright Treaty, art. 11,Dec. 20, 1996, 36 I.L.M. 65, 71; Digital Millennium Copyright Act of 1998, 17 U.S.C. §1201 (Supp. IV 1998).129. See, e.g., Council Directive 01/29/EC of 22 May 2001 on the Harmonisation ofCertain Aspects of Copyright and Related Rights in the Information Society, 2001 O.J.(L 167) 10, art. 6. [hereinafter European Copyright Directive], available athttp://europa.eu.int/eur-lex/pri/en/oj/dat/2001/l_167/l_16720010622en00100019.pdf.The proposed Copyright Directive implementing the WIPO Copyright Treaty contains aprovision prohibiting acts of circumvention and also the manufacturing and selling ofequipment suitable for that purpose. See also Council Directive 91/250/EEC of 14 May1991 on the Legal Protection of Computer Programs, 1991 O.J. (L 122) 42, art. 7(1)(c)(containing earlier recognition of protection of technological means) [hereinafterSoftware Directive], available at http://europa.eu.int/eurlex/en/lif/dat/1991/en_391L0250.html.

competition policy.130It is in this light that the suitability of the UCITA provisionsmust be examined. The debates about what type of intellectualproperty protection is appropriate, particularly in an electronicenvironment, have been occurring on an international as well asdomestic basis. UCITA has a direct effect on that debate, sinceUCITA allows individual licensors to “preempt” that debatethrough licensing.First, of course, it is necessary to determine under existinglaw those rights that cannot be changed by contract. Europeanlegislation grants certain rights to copyright users, for example,that do not depend upon contract for their existence. Four areincluded in the European Software Directive:131 (i) the right tomake a back-up copy;132 (ii) the right to observe, test or study acomputer program;133 (iii) the right to test for error correction;134and (iv) the right to “decompile” or reverse engineer.135 TheSoftware Directive, however, is silent on whether these limits maybe waived or varied by contract. Rights under the EuropeanDatabase Directive include the right to perform acts inherent tonormal usage,136 and the right to re-utilize substantive parts of adatabase.137 Similarly, the law of other countries, in setting out130. The use restrictions and restrictions on assignability are particularlyegregious to some. As one writer has noted:Software publishers expanded copy-use licensing objectives fromprotection against copying to shielding their programs from product-linecompetition, regulating development of interoperable products, andprotecting information content not protected by copyright. Other use andtransfer restrictions are employed to differentiate, and thereby pricediscriminate between, user markets . . . . [and] foreclose the development ofproduct resale or system maintenance competition.Rice, supra note 11, at 281; but see Dessomentet, supra note 104, at 75 (“The veryexistence of a well-developed body of unfair competition law supplementing theunavoidable loopholes of the statutory intellectual property rights prevents Europeansfrom viewing the legislative protection of IP rights as an implicit exclusion of all otherprotection by means of contractual rights.”).131. Software Directive, supra note 129.132. Id. at art. 5(2). That right “may not be prevented by contract insofar as it isnecessary for that use.” Id.133. Id. at arts. 5(3), 9(1).134. Id. at art. 5(1).135. Id. at arts. 6, 9(1).136. Council Directive 96/9/EC of 11 Mar. 1996 on the Legal Protection ofDatabases, 1996 O.J. (L 77) 20, art. 6(1) [hereinafter Database Directive], available athttp://europa.eu.int/index_en.htm.137. Id. at art. 8 (noting that the right may not be overridden).

the intellectual property balance of that regime, may givelicensees of information certain use and other rights that may notbe varied by contract.138 However, an examination of these rights,and the extent to which they are mandatory, is only the beginningof the analysis.139The second and larger question is what the appropriateframework should be for protection of rights in data. As ProfessorRochelle Dreyfus has noted, there is a balance struck in everyintellectual property scheme proposed (copyright, trademark,patent): the right of the creator inventor to exploit and the right ofthe public to use.140 Of course, the balance struck in the UnitedStates is not necessarily the same as the balance struck in othercountries: the inability to reach international consensus onappropriate international intellectual property schemes raises thepossibility of similar difficulties in attempts to reach internationalconsensus on appropriate private contractual schemes controllingthe right to use information. UCITA is one-sided to the extentthat it allows licensors of information to change that balance, i.e.,by allowing a creator of information to tie up that informationlonger, or for more uses, than would be allowed by IP law, withoutrecognition of the public domain and its importance:Each [intellectual property] regime premised is aroundleaks. Each depends upon information being private forlong enough for the creators to earn a profit, but theneach permits the information to leak—drip—into a placewhere the public can use it freely. UCITA, in contrast,finds every hole, and then either plugs it or allows thecontracting parties to plug it. Under UCITA, a licensor138. See, e.g., Murumba, supra note 70, at 107 (discussing such issues underAustralian law and the use of licenses “to destroy the delicate balance between theprivate and public domains that copyright in particular, and intellectual property ingeneral, have maintained over the years”).139. UCITA recognizes that some of its provisions may be preempted by federallaw such as copyright; if federal law invalidates a contract term, then federal lawcontrols. It further recognizes that a license provision may be unenforceable because itcontravenes fundamental public policy “to the extent that the interest in enforcementis clearly outweighed by a public policy against enforcement of the term.” UCITA §105. Provisions such as these avoid, rather than confront, the more critical issues ofwhat those laws or policies should be—which is of course the subject of the ongoingintellectual property law debates. Query whether it is not better to get the balanceright at the outset, rather than to depend upon an external body of law to draw thelimits on enforceability.140. Dreyfuss, Marketplace, supra note 124.

can “take out” a licensee—put a licensee out ofcompetition in the licensor’s creative market forever. Alicensor can bar fair use, reverse engineering, follow-oninventions, and new applications. The licensor canrequire grantbacks, tie ins and tie outs. What is more,UCITA is very efficient. It facilitates mass marketcontracts, and it makes sure those restrictions areimposed on successive licensees and remote purchasers,even when those purchasers are not aware of what theyare letting themselves in for. . . . In the final analysis,what UCITA does is this: in the name of promotingexploitation of this generation of works, it hobbles futuregenerations of creators.141The acceptability of such provisions internationally is far fromclear. As one commentator observed: “from a Europeanperspective, . . . [t]he combination of contract and technology posesa direct threat to the copyright system as we know it, and mayrequire an entierly new body of information law to safeguard thepublic domain.”142 On the other hand, another commentator hasnoted that “most of the European laws’ limitations on copyrightlaws are not mandatory in my view: licensors and licensees canestablish their own regulations for themselves, disregarding thelimitation to the copyright that benefit all people not bound bycontract.”143Whether or not one ultimately agrees with these observations,the point is that adoption of UCITA-like principles on theinternational level without a full inquiry into the merits of sucharguments would be undesirable. Enactment of a uniforminformation transaction code without a “coherent, integralinformation policy”—or at least without consideration of such apolicy—would be dangerous.144141. Id. at 52-53.142. See Hugenholtz, supra note 78, at 78-79 (footnote omitted).143. Dessemontet, supra note 104, at 76.144. Dreyfus, Marketplace, supra note 124; see also Murumba, supra note 70, at 92.Although they effect fundamental changes, these legislative initiatives aresometimes characterized as just an effort to systematize or reconfiguretransactional law into a form useable in the digital environment, but thevigorous worldwide debate which has surrounded these initiatives—especially UCITA—indicates that there is much more at stake here thanmere systematization or adaptation . . . . [What] is at stake is no less thanthe very foundations of an emerging law of the digital domain. In the context

Other Substantive IssuesThere are, of course, controversial provisions within UCITAthat will have special significance on the international level. Forexample, “hidden” within the lengthy act are two provisions ofprofound consequence for all transactions: those addressing choiceof law and choice of forum issues.145 Again, the expansive grant ofparty autonomy in these areas, and whether those rules areappropriate, is the subject of much ongoing debate—both withinthe context of UCITA146 and on a more generalized basis aswell.147 Obviously, before such rules are raised to theof software licensing, that law consists of three interlocking components:contract, intellectual property, and increasingly, the law-like architecturalfeatures and protocols of that domain.Id.145. UCITA §§ 109, 110.146. See, e.g., Amos Shapiro, Territorialism, National Parochialism, Universalismand Party Autonomy: How Does One Square the Choice-of-Law Circle?, 26 Brook. J.Int’l L. 199, 205 (2000) (noting that “a choice-of-law methodology that aspires topromote the idea of objective fair notice ought to refrain from designating as applicablea system of law that clearly has no appropriate relationship with a party to thelitigation”); William J. Woodward, Jr., Contractual Choice of Law: Legislative Choice inan Era of Party Autonomy, 54 SMU L. Rev. 697 (2001). At least one author appears tobelieve that in the case of applicable law, uniformity is nigh impossible. Ted Janger,Jurisdictional Competition and the Dim Prospects for Uniformity, 26 Brook. J. Int’l L.187, 196 (2000).Thus a law of the licensor rule is unlikely to become a uniform choice-of-lawrule under either a decentralized or uniform law approach, and is likely tocreate an incentive towards a substantive race to the bottom if it is adoptedas a mandatory rule of international law. [The] law of the licensee rule wouldnot appear to solve the problem. The same pressures that would likely defeatuniform adoption of a UCITA type rule would work in reverse with regard toa law of the licensee rule.Id.At the 1997 meeting, discussions on choice of law were curtailed when theReporter agreed to reexamine whether the parties to a contract may choose the law of ajurisdiction with no connection to the transaction as the law governing the transaction,as well as the issue of what constitutes a manifestation of assent to the law of a givenjurisdiction. At the 1998 ALI Annual Meeting discussing Article 2B, the membershipappeared to be equally divided over a sense-of-the-house motion to delete thissubsection and the choice of law provisions in their entirety, leaving the issue of theUCC’s contractual choice-of-law provision to be resolved by the Article 1 DraftingCommittee. The decision of the ALI to withdraw from the process precluded anyfurther ALI action on these provisions.147. The changes to the choice of law provisions contained in proposed revisions toArticle 1 of the Uniform Commercial Code have provoked a great deal of discussion.See Woodward, supra note 146; Fred H. Miller, Intrastate Choice of Applicable Law inthe UCC, 54 SMU L. Rev. 525 (2001); Amelia H. Boss, The Jurisdiction of Commercial

international level, it is necessary to examine their consistencywith other law,148 such as the Rome Convention which determineschoice of law149 and the 1968 Brussels Convention on choice offorum,150 and to examine their propriety in light of internationaldiscussions such as those currently taking place in the Hague.151Recognition of party autonomy is taken to an even furtherextreme in UCITA’s condonation of a licensee’s self-help rights,which calls into question the propriety (from both a legal andpolicy perspective) of the use of technological methods to replaceboth contract and intellectual property law as a method ofprotecting a licensor’s rights in information.152Despite the difficulties inherent in adopting UCITA as modelinternational legislation, UCITA is nonetheless useful as a guideto future work in the field. Even the critics of UCITA recognizethe contributions that it has made to the development of theLaw: Party Autonomy in Choosing Applicable Law and Forum under ProposedRevisions to the Uniform Commercial Code, 32 Int’l Law. 1067 (1998); KathleenPatchel, Choice of Law and Software Licenses: A Framework for Discussion, 26 Brook J.Int’l L. 117 (2000).148. One particular aspect of the choice of law rule bears special note: thedifferentiation between the law of jurisdictions inside and outside the United States.Under UCITA, if the laws of a country outside the United States apply as a result ofthe choice of law provisions, that law governs only if it provides substantially similarprotections and rights to a party not located in that jurisdiction as are provided underUCITA. UCITA § 109(c). Significantly, this rule applies only if there is no contractchoice of law provision: the theory being that the parties (including consumers) canagree to give up the protections they would otherwise be entitled to. To the extent thatthis provision demonstrates a “U.S.-centric” provision, its propriety on an internationalbasis is extremely questionable. A similar problem exists in the proposed revisions toArticle 1 of the Uniform Commercial Code.149. Convention on the Law Applicable to Contractual Obligations, opened forsignature June 19, 1980, 1980 O.J. (L 266) 1.150. Convention on Jurisdiction and the Enforcement of Judgments in Civil andCommercial Matters, 1978 O.J. (L 304) 77.151. Hague Conference on Private International Law.152. See generally Robbin Rahman, Electronic Self-help Repossession and You: aComputer Software Vendor’s Guide to Staying out of Jail, 48 Emory L.J. 1477 (1999);Craig Dolly, The Electronic Self-help Provisions of UCITA: A Virtual Repo Man?, 33 J.Marshall L. Rev. 663 (2000). One of the arguments frequently advanced in favor of theself-help provisions is that self-help is already recognized in the Uniform CommercialCode Article 9 provisions dealing with default; this is one area where there isapparently a reluctance to admit that the issues involving electronics are sufficientlydifferent to merit a different rule. See Hugenholtz, supra note 78, at 78-79 (“Thecombination of contract and technology poses a direct threat to the copyright system aswe know it, and may require an entirely new body of information law to guard thepublic domain.”).

law,153 and many of UCITA’s provisions deserve attention andstudy to determine what can be learned from them. UCITA hassucceeded in identifying many of the electronic commerce issuesthat need resolution, even if the solutions it proposes do notultimately prevail. In addition, UCITA has succeeded informulating some more general concepts that represent a stepforward in the development of the law. The challenge is to ferretout the good.One concept that originates in UCITA, that of a “massmarket” license, deserves particular study.154 In the past, muchlegislation has distinguished between transactions based upon theidentity of the parties to the transaction: e.g., merchant versusnon-merchant155 or consumer versus merchant. UCITA, however,recognizes that it may make sense to distinguish betweentransactions based not on the identity of the parties, but on thenature of the contracting process leading to the agreement. Oneshould distinguish the execution from the concept of the massmarket contract: whether or not one agrees with the particulars ofthe definition of the “mass market” license, or with the specifics ofthe rules applicable to such a license,156 the concept may become auseful tool for future analysis.A second example of an attempt in UCITA to deal withlegitimate issues is in the area of mistake. Although mistakesoccur in any environment and with regard to all kinds of subjectmatter, UCITA recognizes that the possibility of mistake in anelectronic environment is sufficiently high and the costs ofprotecting against such error relatively low so that legislativeinducements to reduce error and protection of individuals may beappropriate. As a result, UCITA contains a specific error153. See, e.g., Dreyfuss, Marketplace, supra note 124, at 49 (while criticizingUCITA, expresses enthusiasm for “a system along the lines” of UCITA—“a contractualcode facilitating information exchange, applicable uniformly throughout a tradingregion.”).154. See Holly K. Towle, Mass Market Transactions in the Uniform ComputerInformation Transactions Act, 38 Duq. L. Rev. 371 (2000).155. This is the distinction adopted in Article 2 of the Uniform Commercial Code.156. See, e.g., Chow, supra note 51 (critiquing section 109’s applicability to massmarket licenses). A frequent criticism of UCITA is the failure to recognize that massmarket licenses, to the extent that they are not negotiated and based on true assent,may need special rules in such areas as choice of law and forum and transferabilityrestrictions. See Braucher, supra note 65, at 2.

correction procedure,157 as does the Uniform ElectronicTransactions Act.158 Although there are differences between thetwo formulations,159 they both arguably go beyond the moretraditional rules of mistake found in United States law or inEuropean contract law.160 The need for error correctionprocedures has been acknowledged in the European E-CommerceDirective, which mandates that each member state ensure thatservice providers make available “appropriate, effective andaccessible technical means allowing [the customer] to identify andcorrect input errors, prior to the placing of the order.”161 Thedirective, however, simply mandates the presence of errorcorrection procedures, but gives no remedy in the case ofuncorrected errors. UCITA, as well as the UETA, may helpinform the debate on these issues.162III. THE BIGGER PICTUREHow does the UCITA experience fit into the broaderframework of law making? In many respects, that experience (andits potential international component) raises the question of the157. UCITA § 214.158. UETA § 10.159. UCITA deals with consumers, the UETA with individuals. UCITA applies ifthere has been an error in an electronic message created by a consumer using aninformation processing system, whereas the UETA deals with both errors by theindividual and changes in the message occurring during transmission.160. See Ramberg, supra note 34.161. Council Directive 00/31/EC of 8 June 2000 on Certain Legal Aspects ofInformation Society Services, In Particular Electronic Commerce, In the InternationalMarket (Directive on Electronic Commerce), 2000 O.J. (L 178) 1, art. 11. The directivedoes allow, however, non-consumer parties to agree that such procedures not beprovided, and further exempts from the error correction procedures “contractsconcluded exclusively by exchange of electronic mail or by equivalent individualcommunications.” Thus, it appears to be geared primarily toward web-basedapplications facilitating the placement of orders. Query whether, in the case of amerchant buyer, the Directive would give effect to a provision in a click-wrapagreement on the site waiving any rights to error correction procedures.162. There are, of course, many other concepts in UCITA that merit study, such asits concept of electronic agents. See, e.g., Ian R. Kerr, Spirits in a Material World:Intelligent Agents as Intermediaries in Electronic Commerce, 22 Dalhousie L.J. 188(1999); Jean-Francois Lerouge, The Use of Electronic Agent Questioned underContractual Law: Suggested Solutions on a European and American Level, 18 J.Marshall J. Computer & Info. L. 403 (1999) (noting the need to address issues raised byuse of electronic agents in electronic commerce, particularly in the areas of assent andmistake, but rejecting the UCITA approach and formulation); Ian R. Kerr, Providingfor Autonomous Electronic Devices in the Uniform Electronic Commerce Act, athttp://www.ulcc.ca/en/cls/index.cfm?sec=4&sub=4f.

nature of law-making both domestically and internationally. Atleast one European scholar has characterized the substantialfailure of the states to enact UCITA as a demonstration that theUnited States “has not yet resolved to regulate by law, on anation-wide scale, the basic issues raised by contracting on theNet.”163 Thus, in the United States, common law will develop torespond to the needs of electronic commerce, filling the gaps asneed be. By contrast, the lack of any common law in manyEuropean and other countries means that, if there is to be anybody of law on electronic commerce, it must be imposed by statuteor be “regulatory.” Thus, arguably the need for a detailedstatutory framework for electronic commerce is greater in Europethan in the United States.Yet most of its supporters would undoubtedly balk atcharacterizing UCITA as “regulatory.” UCITA proponents went togreat lengths to assure that, from their perspective, the resultingact was not regulatory and resisted attempts to include provisionsthat they had deemed regulatory in substance.164 This apparentdisagreement as to whether or not UCITA is “regulatory” reveals afar more fundamental difference between UCITA and the type oflegislation one typically finds in civil law or European countries.In the latter, legislation or regulation is most frequently aimed atputting restrictions on what private parties may do, or mandatingwhat they must do. An example is the European Directive onElectronic Commerce, which requires disclosures of prescribedinformation and use of prescribed procedures in commercialtransactions.165 UCITA arguably represents a differentperspective of an appropriate legal framework for electroniccommerce than one might find internationally.166 As one163. Dessemontet, supra note 104, at 61.164. Ring, supra note 64, at 326-27.165. See, e.g., Directive on Electronic Commerce, supra note 161, at art. 5 (requireddisclosures of information by service providers), art. 7 (restrictions on unsolicitedcommercial communications) and art. 11 (mandating procedures to be followed uponreceipt of orders and mandating availability of error correction procedures).166. There are some that might argue that UCITA is an example of “private rentseeking”where private interests influence the lawmaking process to achieveredistributive goals, e.g., licensors of software “dominate the rulemaking process andimpose on the rest of the world a set of rules that diminishes overall welfare but thatmake licensors better off.” Paul B. Stephen, Choice of Law and Its Consequences:Constitutions for International Transactions, 26 Brook. J. Int’l L. 211, 215 (2000); seealso Iain Ramsay, The Politics of Commercial Law, 2001 Wis. L. Rev. 564.

commentator noted:By applying national law rather than some cyberspacelaw that would of course be heavily influenced by U.S.law, we try to keep some autonomy. By applying theconsumer’s law, we apply European law in most caseswhere a U.S. supplier is involved. But for how long canwe check the advance of U.S. notions of e-commerce? Thequestion is moot. Legislative autonomy is importantinasmuch as it allows Europe to bargain for fair termsand conditions when an international convention of somesort will be prepared. Legislative autonomy is not an endin itself, but a means to reach a balanced solution—forcontracting generally, for privacy, for intellectualproperty protection and transactions, and for conflicts oflaw.167Despite this apparent conflict, there is a lot to be learned fromUCITA and the drafting experience. One view is that the processis just beginning. There may be a lesson to learn here from theUCC; it was initially completed by the National Conference andthe ALI in 1952, and Pennsylvania quickly became the first toenact it. Adoptions stopped, however, when the Code came underintense scrutiny and attack, leading the drafters to abandonenactment efforts and completely overhaul the Code, resulting ina finally enactable version in 1958.168 It may well be time to sitback and reflect on UCITA: its contributions, its controversies andits criticisms.169 It is time for the international community tocritically scrutinize and (where appropriate) adopt and adaptprovisions from UCITA, allowing a symbiotic process to occurwhich furthers the improvement and harmonization of law.170167. Dessemontet, supra note 104, at 64.168. For an interesting comparison of the drafting histories of both the UCC andUCITA, see Thomas J. Murphy, It’s Just Another Little Bit of History Repeating:UCITA in the Evolving Age of Information, 30 Golden Gate U. L. Rev. 559 (2000).169. See O’Rourke, supra note 73, at 657 (“The point is that [UCITA] should not beabandoned. Rather, it should provide the basis for a national and internationaldiscussion of how best to reconcile the sometimes conflicting goals of flexibility anduniformity in the context of an overarching desire to encourage global electroniccommerce.”).170. Amelia H. Boss, Electronic Commerce and the Symbiotic Relationship BetweenInternational and Domestic Law Reform, 72 Tul. L. Rev. 1931 (1998) (analyzing thesymbiotic process in the area of electronic signatures and electronic records leading tothe UNCITRAL Model Law on Electronic Commerce and the Uniform Electronic

Transactions Act).

Notes and CommentsMaking Waves In Statutory SafeHarbors: Reevaluating InternetService Providers’ Liability for Third-Party Content and CopyrightInfringementWith the growth of the Internet1 as a medium ofcommunication, the appropriate standard of liability for accessproviders has emerged as an international legal debate.2 InternetService Providers (ISPs)3 have faced, and continue to face,potential liability for the acts of individuals using their services toaccess, post, or download information.4 Two areas of law wherethis potential liability has been discussed extensively aredefamation and copyright law.5In the United States, ISP liability for the content of thirdpartypostings has been dramatically limited by theCommunications Decency Act of 1996 (CDA).6 However, ISPs inAmerica do not face the same degree of immunity for third-party1. See generally Reno v. Am. Civil Liberties Union, 521 U.S. 844, 849-51 (1997)(describing the development and operation of the Internet).2. See generally Rosa Julia-Barcelo, Liability for On-Line Intermediaries: AEuropean Perspective, 1998 Eur. Intell. Prop. Rev. 20(12) 453-63 (describing the issueof ISP liability and its dimensions).3. See Mitchell P. Goldstein, Service Provider Liability for Acts Committed byUsers: What You Don’t Know Can Hurt You, 18 J. Marshall J. Computer & Info. L. 591,592 n.1 (2000) (describing Internet service providers as computer systems providingindividuals with modem access to the Internet).4. See Julia-Barcelo, supra note 2, at 453.5. See Religious Tech. Ctr. v. Netcom On-Line Communication Services, Inc., 907F. Supp. 1361, 1367 n.10 (N.D. Cal. 1995).6. 47 U.S.C. § 230 (1996).237

acts of copyright infringement.7 Under the Online CopyrightInfringement Liability Limitation Act (OCILLA), Title II of theDigital Millennium Copyright Act (DMCA),8 Internet serviceproviders are protected from strict liability for third-party acts ofcopyright infringement only if they can prove that their activitiesfit into one of four general safe harbor categories: transitorynetwork communications, system caching, information stored onnetworks under user direction, and information location tools.9 Toretain this safe harbor immunity, OCILLA also requires an ISP topromptly remove any allegedly infringing materials stored on itsservice once it has been notified of their unauthorized use by acopyright holder.10In contrast to the United States, the European Union (EU) isin the process of developing a unified standard for determining theextent of ISP liability for the acts of their third-party users.11 Thenew EU E-Commerce Directive12 addresses these issues, andcreates safe harbor provisions for ISPs that are, for the mostpart,13 very similar to those enforced in the United States by theDMCA.14 The E-Commerce Directive and the DMCA differ in thescope of their liability limitations. The Directive is designed toregulate all instances of ISP liability for the acts of its third-partyusers, and applies one standard horizontally to all areas of law,15including defamation and copyright infringement. This is unlikethe vertical approach to ISP liability taken in the United States,where liability limitations for Internet providers are definedaccording to the area of law in which the offense occurs.16 While7. See Goldstein, supra note 3, at 638.8. 17 U.S.C. §§ 1201-1332 (Supp. IV 1998). Title II of the DMCA is entitled theOnline Copyright Infringement Act. See 17 U.S.C. § 512 (Supp. IV 1998).9. See 17 U.S.C. § 512(a)-(d), (f), (g), (i).10. See id.11. See Rosa Julia-Barcelo, On-Line Intermediary Liability Issues: ComparingE.U. and U.S. Legal Frameworks, 2000 E.I.P.R. 22(3) 105-19.12. See Proposal for a Directive on Certain Legal Aspects of Electronic Commercein the Internal Market, COM (98) 586 final; Directive on Certain Legal Aspects ofInformation Society Services, in Particular Electronic Commerce, in the InternalMarket, Council Directive 00/31, 2000 O.J. (L 178) [hereinafter EU E-CommerceDirective].13. The EU E-Commerce Directive does not include a safe harbor for informationlocation tools, which is present in the DMCA. See id.; see also Julia-Barcelo, supra note11, at 114.14. See Julia-Barcelo supra note 11, at 108.15. See id.16. See id.

the safe harbor provisions limiting ISP liability for user acts ofcopyright infringement will be similar in the United States andthe EU following the adoption of the Directive,17 the safe harborprovisions limiting ISP liability for third-party content will differdramatically.18 By applying the safe harbors of the E-CommerceDirective horizontally, the EU will look to these same provisionsin determining when an ISP can face liability as the publisher ordistributor of a user’s defamatory statement.19 In the UnitedStates, an ISP holds blanket immunity from being considered thepublisher of a third-party defamatory statement under the CDA.20This paper will first discuss the development of the CDA andits requisite safe harbor provision limiting the liability of ISPs fordefamatory content on their systems posted by third parties.21 Itwill next discuss the liability of ISPs for third-party copyrightinfringement under the safe harbor provisions provided by theDMCA in OCILLA.22 It will then compare the liability limitationsof the DMCA to those included in the European Union E-Commerce Directive,23 and outline the implications of applyingthe same safe harbor limitations horizontally to include ISPliability under defamation law within the same safe harborparameters. This paper will conclude that the DMCA’s safeharbor provisions limiting ISP copyright liability for third-partyinfringement as outlined should be applied horizontally indefamation cases to determine whether an ISP can be held liableas the publisher or distributor of a third-party defamatorystatement posted on its service. Finally, this paper will considerthe First Amendment concerns arising from the required removalprovisions of the DMCA,24 and how these provisions could belimited in a defamation context so as not to improperly impede17. See generally id. (comparing the EU E-Commerce Directive to the safe harborprovisions of the DMCA and concluding that they are largely similar).18. See id. at 108.19. See id.20. See Goldstein, supra note 3, at 638.21. 47 U.S.C. § 230 (1996).22. 17 U.S.C. §§ 1201-1332 (Supp. IV 1998).23. See Council Directive 2000/31/EC 2000 O.J. (L 178); see also Julia-Barcelo,supra note 2.24. See generally Alfred C. Yen, Internet Service Provider Liability for SubscriberCopyright Infringement, Enterprise Liability, and the First Amendment, 88 Geo. L.J.1833 (2000) (discussing the First Amendment’s under-emphasized implications oncopyright law).

these rights.I. ISP LIABILITY FOR THIRD-PARTY ACTS OF DEFAMATIONAt common law one who “repeats” the statements of anotheris just as responsible for their defamatory content as the originalspeaker.25 Thus, the publisher of a defamatory statement isequally liable to the defamed party as is the person originallymaking the statement. However, newsvendors, bookstores,libraries, and other vendors and distributors of defamatorypublications are not liable if they did not know or have reason toknow of the defamation.26 Freedom of speech and the pressprevent the government from imposing strict liability ondistributors for the content of materials they carry.27 Therefore,in the quest to avoid liability for the content placed on theInternet by third-party subscribers, it is advantageous under acommon law regime for an ISP to be characterized as a distributorrather than a publisher of an allegedly defamatory statement.A. ISP Liability for Third-Party User Defamation Before the CDAThe first time that the question of liability for an ISP arose inan American defamation case was in Cubby, Inc. v. CompuServe,Inc.28 In this case, CompuServe, an Internet service provider,operated a bulletin board that contained a daily newsletterpublication catering to the journalistic community entitled“Rumorville.”29 Cubby, Inc. began a competing electronicpublication entitled “Skuttlebut.”30 Following the publication ofSkuttlebut, the Rumorville newsletter published statementsclaiming Skuttlebut was publishing information that it obtainedthrough a “back door” that had first been published by Rumorville,that its principal had been “bounced” from his prior job, andcharacterizing Skuttlebut as a “start-up scam.”31 Cubby thensued CompuServe as the publisher of these defamatorystatements.25. See Cinanci v. New Times Pub. Co., 639 F.2d 54, 60-61 (2d Cir. 1980).26. See Lerman v. Chuckleberry Publ’g, Inc., 521 F. Supp. 228 (S.D.N.Y. 1981).27. See Smith v. California, 361 U.S. 147, 152-53 (1959).28. See Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 139 (S.D.N.Y. 1991).29. See id. at 135.30. See id.31. See id.

The court held that CompuServe acted as a distributor of theRumorville newsletter rather than its publisher.32 This holdingwas based on the court’s finding that CompuServe did not exerciseeditorial control over the Journalism Forum bulletin board, butrather had contracted with an independent company, CameronCommunications, Inc., to “manage, review, create, delete, edit andotherwise control [its] contents” in accordance with the editorialand technical standards of CompuServe.33 Further, the courtfound that Rumorville was published by Don FitzpatrickAssociates, who accepted total responsibility for its contents.34CompuServe’s only activity in managing the bulletin board wasrequiring Cameron Communications to limit access to Rumorvilleto those CompuServe subscribers having a membershiparrangement with Don Fitzpatrick.35 Since CompuServe had “noopportunity to review” the content of the newsletter before it wasuploaded and made available to subscribers, the court found thatCompuServe exercised no more editorial control over Rumorville“than does a public library, bookstore or newsstand” over materialit makes available to the public.36 The court went on to hold thatthe proper standard of liability to be applied to CompuServe inthis case was whether it, as a distributor of Rumorville, “knew orhad reason to know of the allegedly defamatory Rumorvillestatements.”37 Since Cubby offered no evidence that CompuServeknew or had reason to know of the statements at issue, the courtrefused to hold CompuServe liable for their content.38Cubby’s holding was not followed by a later case, StrattonOakmont, Inc. v. Prodigy Services Co.39 In this case, ananonymous user of Prodigy’s “Money Talk” bulletin board postedstatements claiming that Stratton Oakmont, a securitiesinvestment banking firm, and Daniel Porush, Stratton’s president,committed criminal and fraudulent acts and was a “cult of brokerswho either lie for a living or get fired.”40 Unlike, the decision in32. See id. at 140.33. Id.34. See Cubby, 776 F. Supp. at 137.35. See id.36. Id.37. Id. at 140-41.38. See id. at 141.39. See Stratton Oakmont, Inc. v. Prodigy Servs., Co., 1995 WL 323710 at *1 (N.Y.Sup. Ct. May 24, 1995).40. Id.

Cubby, the court in Stratton Oakmont held that Prodigy could beheld liable for libel as if it were a newspaper or broadcasterbecause it reserved the right to exercise editorial control over itselectronic bulletin boards.41 The court reasoned that Prodigy helditself out to the public as an ISP exercising editorial control overits bulletin board postings.42 Prodigy employed both an automaticscreening software program and volunteer bulletin board leadersto supervise the content of board submissions.43 The StrattonOakmont court likened these features to the editorial controlexercised by a newspaper, and held that Prodigy was a publisherof the defamatory statements at issue, even though the ISP had noknowledge of these statements.44The court in Stratton Oakmont distinguished the Cubbydecision by describing the different roles performed byCompuServe and Prodigy, the two ISPs and bulletin boardoperators in question. The court acknowledged that a meredistributor, such as a bookstore or library, may be liable for adefamatory statement made by a third party only if it knew or hadreason to know of the defamatory statement at issue.45 Adistributor, such as CompuServe in Cubby, acts as a passiveconduit and will not be found liable for libel in absence of fault. 46The court reasoned that Prodigy was not such a passive conduitbecause it exercised editorial control over its bulletin boardpostings equivalent to that of a newspaper editor.47 By exercisingthis editorial control, the court found that Prodigy exposed itself togreater liability than other computer networks choosing not tomonitor the content of their system.48B. Congressional Reaction to Stratton Oakmont v. ProdigyServices Co.: The CDAIn response to the decision in Stratton Oakmont, Congresspassed Section 230 of the Consumer Decency Act of 1996.49 In41. See id. at *5.42. See id. at *2.43. See id. at *4.44. See Stratton Oakmont, 1995 WL 323710 at *3.45. See id. at *3.46. See id.47. See id.48. See id. at *5.49. 47 U.S.C. § 230 (1996).

this section, Congress provided that online service providers andweb site owners be regarded as distributors rather thanpublishers of third-party content posted on their systems.50Congress adopted this section with the intention of removing theheightened liability Stratton Oakmont placed on ISPs who chooseto monitor the content posted on their systems that was generatedby other, third-party providers.51 Section 230 also limits theliability of an online provider for any action taken voluntarily bythe ISP in “good faith” to restrict access to material “that theprovider or user considers to be obscene, lewd, lascivious, filthy,excessively violent, harassing, or otherwise objectionable, whetheror not such material is constitutionally protected.”52 Theseprovisions were enacted to facilitate Congress’ larger goal of“removing disincentives for the development and utilization ofblocking and filtering technologies” which allow parents to“restrict their children’s access to objectionable or inappropriateonline material.”53Although portions of the CDA prohibiting the transmission of“obscene or indecent” communications were held unconstitutionalby the Supreme Court in Reno v. American Civil LibertiesUnion,54 the provisions limiting ISP liability for third-partycontent remain viable.55 While the language of Section 230clearly limits the liability ISPs face as publishers of third-partycontent, it does not directly address the remaining liability thatISPs could potentially incur as distributors of third-partydefamation when the required knowledge standard is proven incourt.56 However, in applying the CDA to online defamationcases, the courts have declined to draw a distinction betweendistributors and publishers in a way that would give rise to ISP50. 47 U.S.C § 230(c)(1) provides: “No provider or user of an interactive computerservice shall be treated as the publisher or speaker of any information provided byanother information content provider.”51. See Joint Explanatory Statement of the Congressional Conference Committee,H.R. Conf. Rep. No. 104-458 (1996), reprinted in 1996 U.S.C.C.A.N. 124.52. 47 U.S.C. § 230(c)(2)(A) (1996).53. 47 U.S.C. § 230(b)(4) (1996).54. 521 U.S. 844 (1997).55. See id; see also Goldstein, supra note 3, 617-21 (discussing the limitations ofthe Reno case on the CDA and the continuing viability of Section 230 as it limits thecontent liability of ISPs).56. See Joseph P. Zammit & Felicia Gross, Web Site Liability: Risks and Costs ofCompliance, 4 P.L.I. Ann. Internet L. Inst., Course Handbook Series, available at 611PLI/Pat 815, 895 (2000).

liability for third-party content.57 In fact, as this paper will nowdiscuss, courts have interpreted the CDA as providing ISPs acomplete safe harbor from tort liability for third-party contentposted on their services, even when the provider does not employany of the self-monitoring activities Congress hoped to promote inenacting the CDA.58C. Applying the CDA: ISP Immunity from Liability for Third-Party ContentIn Zeran v. America Online, Inc.,59 an unidentified personposted a message on an America Online bulletin board advertisingt-shirts for sale glorifying the April 16, 1995 bombing of the AlfredP. Murrah Federal Building in Oklahoma City, and instructinginterested parties to call a person named “Ken” at Zeran’s homephone number.60 After receiving a high volume of calls, Zeraninformed America Online of the defamatory posting, and was toldthat the message would be removed.61 Zeran later sued AOL forunreasonable delay in removing the message, refusing to post aretraction, and failing to screen for similar postings following theincident.62 AOL successfully plead that it was immune fromliability under the CDA.63On appeal, the court construed Section 230 of the CDAbroadly, stating that “[b]y it’s plain language, Section 230 createsa federal immunity to any cause of action that would make serviceproviders liable for information originating with a third-party userof the service.”64 The court also noted that the CDA directlyprecluded an ISP such as America Online being liable as thepublisher of third-party content.65 The court regarded AOL’sactivities in maintaining the bulletin board not as the role of a57. See id; see also Douglas B. Luffman, Defamation Liability for Online Services:The Sky is Not Falling, 65 Geo. Wash. L. Rev. 1071, 1083-85 (1997); David R. Sheridan,Zeran v. AOL and the Effect of Section 230 of the Communications Decency Act UponLiability for Defamation on the Internet, 61 Alb. L. Rev. 147, 167-77 (1997).58. See e.g.,Blumenthal v. Drudge, 992 F. Supp. 44, 52 (D.D.C. 1998); Zeran v.Am. Online, Inc., 129 F.3d 327, 330-31 (4th Cir. 1997), cert. denied, 524 U.S. 937(1998).59. See Zeran, 129 F.3d at 327.60. See id. at 328.61. See id.62. See id.63. See id.64. Id. at 330.65. See Zeran, 129 F.3d at 330.

distributor, but rather as a participant in the publication of thirdparty content—an activity which is protected under Section 230.66The court even went so far as to hold that, upon notice of adefamatory posting, an ISP is “thrust into the role of a traditionalpublisher,” because the service provider must then determinewhether or not to publish, edit, or withdraw the posting.67 Thisrationale effectively prevents an ISP distributing third-partycontent from ever meeting the knowledge standard required forliability, because once knowledge is proven, the ISP would rise tothe federally protected realm of a publisher.68 In support of thisholding, the court argued that an interpretation of Section 230imposing liability on ISPs with knowledge of defamatory contentwould be contrary to Congress’ intent in passing the CDA, as itwould deter service providers from monitoring the content of theirsystems.69One year after Zeran, a similar decision was reached inBlumenthal v. Drudge70 regarding the CDA’s extreme limitationon the liability faced by an ISP for third-party content posted onits services. In Blumenthal, Sidney Blumenthal, an assistant toPresident Clinton, sued Matt Drudge, a gossip columnist, overcomments included in Drudge’s publication, “The Drudge Report,”suggesting that Blumenthal had a history of beating his wife.71Blumenthal also sued America Online, which had a contract withDrudge to make the Drudge Report available to all AOL membersfor the period of one year in exchange for a monthly royalty of$3,000.72 In the contract, America Online reserved the right toremove content from the Drudge Report that it reasonablydetermined violated AOL’s standard terms of provider service.73AOL had also vigorously promoted its distribution of the DrudgeReport, characterizing Matt Drudge as a “Runaway GossipSuccess.”74 After Drudge formally retracted his allegations aboutBlumenthal, America Online disabled access to the story on its66. See id. at 332.67. See id.68. See Goldstein, supra note 3.69. See Zeran, 129 F.3d at 333; see also Goldstein, supra note 3.70. 992 F. Supp. 44 (D.D.C. 1998).71. See id.72. See Blumenthal v. Drudge, 992 F. Supp. 44, 47 (D.D.C. 1998).73. See id.74. See id. at 51.

network and removed the edition of the Drudge Report from itsonline archives.75 In Blumenthal, the court held that Section 230of the CDA prevented AOL from being liable in tort as a publisheror distributor of the Drudge Report:AOL is not a passive conduit like the telephonecompany, a common carrier with no control and thereforeno responsibility for what is said over the telephonewires. Because it has the fight [sic] to exercise editorialcontrol over those with whom it contracts and whosewords it disseminates, it would seem only fair to holdAOL to the liability standards applied to a publisher or,at least, like a book store owner or library, to the liabilitystandards applied to a distributor. But Congress hasmade a different policy choice by providing immunityeven where the interactive service provider has an active,even aggressive role in making available contentprepared by others . . . Congress has conferred immunityfrom tort liability as an incentive to Internet serviceproviders to self-police the Internet for obscenity andother offensive material, even where the self-policing isunsuccessful or not even attempted.76In reaching this decision, Blumenthal embraced Zeran’sconclusion that the CDA immunizes ISPs from content-based tortliability whether or not the provider can be shown to have actualor constructive knowledge of improper content.77The blanket immunity the CDA furnishes Internet serviceproviders for third-party content is problematic. Under theholdings of Zeran and Blumenthal, an ISP can escape liability forthird-party content completely, without an evaluation of theprovider’s knowledge as to the content of the material, or an indepthanalysis of the provider’s role in transmitting the material.In Blumenthal, America Online had formally contracted withDrudge to distribute his Drudge Report and actively promoted thenewsletter as a gossip success. Rather than examining thispreexisting relationship between Drudge and AOL to determinethe extent of the ISP’s knowledge as to the content of the DrudgeReport, the court awarded complete immunity to America Online75. See id. at 47.76. See Blumenthal, 992 F. Supp. at 51-52 (footnotes omitted).77. See Blumenthal, 992 F. Supp. at 50 (quoting Zeran, 129 F.3d at 330-31).

due to its status as an Internet service provider. By establishingthat an ISP will never be considered a publisher of third-partycontent, the CDA prevents courts from fully analyzing the natureand extent of the provider’s communications to determine liability.As a result, the CDA establishes content immunity for all ISPswho host third-party postings on the Internet, regardless of howthis information is transmitted, stored, or accessed by otherusers.78 An analysis of how the ISP maintains the contentinformation on its service is potentially important in determiningthe extent of the ISP’s knowledge.The importance of these principles is illustrated by anotherdefamation case following the enactment of the CDA. In Lunneyv. Prodigy Services Co.,79 a teenage Boy Scout, Lunney, suedProdigy claiming that he had been stigmatized and defamed by athird-party user who opened a number of Prodigy accounts underLunney’s name and posted vulgar messages on Prodigy’s bulletinboards, attributing them to Lunney.80 Prodigy monitored thecontent of these bulletin boards, and did in fact close down severalof the Lunney accounts due to obscene postings and inaccurateuser profile information.81 These events took place prior toCongress’ enactment of the CDA, but the decision was renderedafter the CDA became law.82 Without retroactively applying theimmunity provisions of the CDA, the New York Court of Appealsheld that Prodigy was not the publisher of the statements at issue,but rather served the function of a conduit, transmitting thirdpartymaterial without exercising editorial control over itscontents.83 The court likened Prodigy’s role in transmitting e-mailmessages to that of a telephone company connecting individualspeakers through electronic means, and concluded that it is notexpected that either conduit will monitor the content of its78. Although immunity for third-party content is established through Section 230for ISPs, some courts are extending this statutory immunity to other providers ofInternet services, such as online department stores. See Schneider v. Amazon.com,Inc., 31 P.3d 37 (Wash. Ct. App. 2001) (affirming lower court’s holding thatAmazon.com was a protected publisher of third-party content under Section 230 as faras its customer book reviews were concerned).79. 723 N.E.2d 539 (N.Y. 1999), cert. denied, 529 U.S. 1098 (2000).80. See id. at 539-40.81. See id. at 540.82. See id. at 543.83. See id.

subscribers’ conversations.84 Further, the court held thatProdigy’s independent acts of monitoring bulletin board postingsdid not alter its passive character in providing Internet service.85Thus, the court held that Prodigy, as a mere conduit fortransmitting third-party data, was not a publisher or distributorof the material for the purposes of tort law, and found no liabilityfor the ISP.86While Lunney’s holding is similar to Zeran and Blumenthal inthat the court found that an ISP was not liable as the publisher ordistributor of third-party data, Lunney differs in its analysis.Rather than regarding Prodigy’s actions in transmitting thirdpartye-mail messages to its bulletin board as a publication andthen resorting to the congressional safe harbor extinguishing ISPliability for third-party content, Lunney looked instead to theactual function performed by Prodigy in transmitting thesemessages. In concluding that Prodigy acted as a mere conduittransmitting the independent content of its third-party users, theLunney court evaluated the function performed by the ISP in thatcase. It then used this finding to determine both the ISP’s role aseither a publisher or distributor of the third-party content and theprovider’s knowledge of the material’s content. By regarding theISP as a medium for transmitting data rather than a form ofpublishing or distribution, the Lunney court came closer to fullyevaluating the role of an ISP as a medium for communication.Lunney’s holding suggests that the blanket immunityprovisions of the CDA are not necessary for a court to hold that anISP is not liable as the publisher or distributor of third-partycontent posted to its Internet service. Rather than resorting tocomplete immunity under the CDA, a more appropriate schemefor a court determining ISP content liability would constitutemaking a factual finding as to the role of the ISP in presenting thematerial to the public and then using this finding to determine theextent of the provider’s knowledge and control of the third-partycontent. This line of analysis would provide a technique for courtsto evaluate the extent of ISP liability in cases like Blumenthal,where the role of the ISP more resembles a publisher, withoutresorting to a blanket foreclosure on provider liability.84. See id. at 541-42.85. See Lunney, 723 N.E.2d at 542.86. See id.

There are several different factors that are relevant indetermining the role of an ISP in presenting third-party content tothe public. A court could examine the actions of the individualprovider in light of traditional defamation law, and attempt todetermine liability based on whether the provider acts more like apublisher or a distributor of the statement at issue. Anotherpossibility would be to compare an ISP with another form ofconduit, like a telephone company.87 This comparison is not allinclusive,however, because it covers only one of several possibleways that an ISP could function to transmit information to thepublic, some which could arguably involve greater knowledge onthe part of the provider than others. The best comparison, as thispaper will now suggest, is to the provisions in the DMCA limitingISP liability for third-party copyright infringements. If theselimitations were applied to defamation law in the United States asthey are in Europe, ISPs would face uniform liability for thirdpartyacts under OCILLA’s transmission-specific safe harbors.These specific provisions would also properly limit immunity tothose functions of ISPs that truly serve an innocuous function, inthe realms of both copyright and defamation law.II. ISP LIABILITY FOR THIRD-PARTY ACTS OF COPYRIGHTINFRINGEMENTCopyright owners hold, among other rights, an exclusive rightto control the reproduction, distribution, display, or performanceof their protected material.88 The U.S. Copyright Act providesthat “[a]nyone who violates any of the exclusive rights of thecopyright owners . . . is an infringer . . .”89 Subject to therecognized fair use exception, any reproduction of a protectedwork constitutes copyright infringement, and can create liabilityfor the infringer.90 Three standards of liability for copyrightinfringement have developed over the years and have beendiscussed with regard to ISPs.91 Direct liability ensues when anindividual or entity violates any exclusive right of a copyright87. See id.88. 17 U.S.C. § 106 (1996), amended by Pub. L. No. 104-39, 109 Stat. 336.89. 17 U.S.C. § 501(a).90. See Bruce P. Keller, Internet and Online Law: Copyright, 4 P.L.I. Ann.Internet L. Inst., Course Handbook Series, available at 610 PLI/Pat 169, 201 (2000).91. See id.

holder.92 Contributory liability applies to any person who, withknowledge of the infringing activity, “induces, causes or materiallycontributes to the infringing conduct of another.”93 Vicariousliability has been held to exist when the subject of a copyrightinfringement suit both exercised the right and ability to supervisethe primary infringer and had a direct financial interest in theexploitation of copyrighted materials.94Direct and vicarious copyright infringers are subject to strictliability for their offenses, regardless of their intent or knowledgeat the time of the violation.95 Even innocent copyrightinfringements are subject to strict liability under these twostandards.96 Contributory infringement, however, depends on thecontributor’s actual or constructive knowledge that a copyrightviolation is taking place.97 In assessing the appropriate level ofliability facing an ISP in a particular factual situation, one mustlook at the provider’s actions to determine whether the role of theprovider is one of direct involvement in the unauthorized use. Ifthe provider’s role is not direct, then one must consider whetherthe provider has sufficient knowledge and involvement to becontributorily liable or, alternatively, whether the provider exertsenough control and derives financial benefit from the infringementso as to be vicariously liable.98A. Copyright Liability of ISPs Before the DMCAThe first case in the United States dealing with the subject ofonline service provider liability for third-party copyrightinfringement was Playboy Enterprises v. Frena.99 Here, Playboysued George Frena, the operator of an online bulletin boardservice, for copyright infringement, claiming that copyrightedPlayboy photographs had been posted on Frena’s service without92. See id.93. Gershwin Publ’g Corp. v. Columbia Artists Mgmt., Inc., 443 F.2d 1159, 1162(2d Cir. 1971). See also Yen, supra note 24, at 1843-44.94. See Gershwin Publ’g, 443 F.2d at 1162; see also Yen, supra note 24, at 1843-44.95. See Irina Y. Dmitrieva, I Know It When I See It: Should Internet ProvidersRecognize Copyright Violation When They See It?, 16 Santa Clara Computer & HighTech. L.J. 233, 235 (2000).96. See id. at 236.97. See id.98. See Keller, supra note 90, at 241.99. 839 F. Supp. 1552 (M.D. Fla. 1993).

authorization.100 Frena defended himself by claiming that thecopyrighted pictures had been uploaded to his bulletin board byindividual subscribers without his knowledge, and that heremoved the photographs as soon as he received notification fromPlayboy.101 The court found that the pictures in question were infact the copyrighted property of Playboy, and that in someinstances these copyrighted photographs had been displayed in analtered form on Frena’s service, such that Playboy’s trademarkswere replaced with Frena’s own advertisements.102 From thispromotional alteration, the court concluded that Frena hadproduced a “product” containing infringing material, and haddirectly violated both the distribution and display rights that werereserved to Playboy.103 Further, the court disregarded Frena’slack of knowledge defense, holding Frena strictly liable for theinfringement.104Two years after Frena was decided, another court evaluatedthe liability of bulletin board operators and Internet serviceproviders for third-party acts of infringement. In ReligiousTechnology Center v. Netcom On-Line Communication Services,Inc.,105 the Church of Scientology sued Dennis Erlich, a formerscientologist minister, for his unauthorized posting of copyrightedchurch materials on a Usenet bulletin board.106 Also joined in thesuit were Klemesrud, the bulletin board operator managing theboard Erlich used to display the messages, and Netcom, Inc., theISP connecting Klemesrud’s bulletin board to the Internet.107 TheChurch joined Netcom and Klemesrud after it notified both partiesof the infringing activities, and both the ISP and the bulletinboard operator refused to remove the postings without a showingof proof that they actually contained infringing material.108 Afterreviewing both the Frena decision and other persuasive caselaw,109 the court granted summary judgment in favor of Netcom100. See id. at 1554.101. See id.102. See id. at 1556.103. See id.104. See Frena, 839 F. Supp. at 1558.105. 907 F. Supp. 1361 (N.D. Cal. 1995).106. See id. at 1365.107. See id.108. See id. at 1366.109. See id. at 1371 (discussing Sega Enterprises Ltd. v. MAPHIA, 857 F. Supp.679 (N.D. Cal. 1994)).

and Klemesrud on the counts of direct and vicarious liability forthe alleged copyright infringement, but sustained the contributoryliability cause of action against both parties.110In Netcom, the court concluded that Frena addressed adifferent factual situation than the claims advanced by theChurch of Scientology. In Frena, Playboy focused its case on theunauthorized display and distribution of its protectedphotographs, and the court found that Frena had produced a“product” violating these rights.111 The situation was different inNetcom, where the Church claimed that Netcom and Klemesrudengaged in illegal reproduction of copyrighted material. Here, thecourt held that the mere “storage on a defendant’s system ofinfringing copies and retransmission to other services” does notconstitute direct infringement by a bulletin board operator whenthe infringing information was originally uploaded by a thirdpartyuser.112 Similarly, the Netcom court held that an ISP, likeNetcom, could not face direct liability for third-party copyrightinfringement when it performs only the function of a conduit, anddoes not hold files in an archive for long periods of time.113 Thecourt also found that while the Church did allege a genuine issueof fact regarding Netcom’s control over the postings placed on itsbulletin boards, the Church did not offer significant proof that theISP incurred a financial benefit from the infringing activities.114Thus, the court did not sustain the Church’s claim against Netcomfor vicarious liability.115The Netcom court did, however, refuse to grant the motionsagainst both the ISP and Klemesrud for summary judgment onthe count of contributory liability.116 The court concluded that agenuine issue of fact existed as to whether the defendants hadactual knowledge of the infringing activity. Since the Churchnotified Netcom and Klemesrud of the misused material before theposting was set to expire, the two defendants’ failure to removethe infringing statements could be construed as contributing to a110. See Netcom, 907 F. Supp. at 1383.111. See Frena, 839 F. Supp. at 1556.112. See Netcom, 907 F. Supp. at 1371.113. See id. at 1372.114. See id. at 1375, 1376-77.115. See id.116. See id. at 1374-75.

known infringement.117 Thus, the court allowed this claim toproceed to trial.118Unlike Frena, the court in Netcom held that an Internetservice provider serving only the function of a conduit for thirdpartyinformation cannot be held liable for direct copyrightinfringement, because the actual reproduction was made at thehands of the independent third party rather than the ISP.Although Frena dealt with the direct copyright liability of abulletin board operator for alleged third-party postings, aninference can be made that the same liability would attach to anISP in a similar situation. Netcom rejected this notion, likeningan ISP to a telephone company or other communications conduit,which exercises no editorial control over third-party submissions,but instead merely transmits them innocuously.119 Thus,Netcom’s line of reasoning implies that an ISP cannot be heldliable for direct copyright infringement when the infringingactivity originates with a third-party user. As a result, Netcomand Frena suggest conflicting standards of copyright liability forInternet service providers.B. Legislating ISP Copyright Liability: the DMCAFollowing the decisions in Frena and Netcom, scholarlyopinion was divided regarding the proper standard of copyrightliability that should, as a rule, be imposed on ISPs for theiractivities as service providers to third parties.120 As mentionedpreviously, holding ISPs either directly or vicariously liable forthird party copyright infringement would result in strict liability,while holding them to a contributory liability standard, as thecourt did in Netcom, bases liability of the ISP’s awareness of theinfringing activity.121 This debate was also an issue in Congress.In the early 1990s, the Clinton administration had organized atask force on the National Information Infrastructure, whosepurpose was reworking copyright laws to make them conform withthe new realities of the emerging digital society.122 This group’s117. See id. at 1375.118. See Netcom, 907 F. Supp. at 1375.119. See id. at 1370.120. For an overview of some of the different scholarly approaches to this questionsee Dmitrieva, supra note 95, at 237.121. See text supra; see also Dmitrieva, supra note 95, at 237.122. Info. Infrastructure Task Force, Intellectual Property and the National

eport, the White Paper on Intellectual Property and the NationalInformation Infrastructure,123 recommended that ISPs be treatedas distributors under copyright law, and that they should be heldto a strict liability standard for their users’ acts ofinfringement.124 Further, the report did not recommend anymodifications in copyright law to limit ISP liability.125The White Paper met with resistance from ISPs, who arguedthat such a strict liability standard would expose them tounreasonable liability for third-party acts, and would require ISPsto monitor their systems.126 ISPs also claimed that it would beimpossible to catch all copyright violations through monitoring,127and that extensive monitoring would produce a chilling effect onfree speech.128 Internet providers preferred an actual knowledgestandard, whereby they would face liability for third-partycopyright infringement only in cases where a provider knew of theinfringement and took no action to remove it promptly.129 Incontrast, copyright owners argued that a strict liability standardwas appropriate for ISPs, and was the only way to adequatelyprotect owners’ rights.130 Copyright holders directly opposed anactual knowledge standard, arguing that it would remove anyincentive for an ISP to monitor its system, rewarding ignorancewith limited liability.131While the Congressional debate over copyright liability forISPs continued, the United States became a party to two WorldIntellectual Property Organization (WIPO)132 Treaties, created inInformation Infrastructure: The Report of the Working Group on Intellectual PropertyRights (Bruce Lehman ed., 1995).123. See id.124. See id. at 212.125. See id.126. See NII Copyright Protection Act of 1995: Hearing on S.1284 Before the SenateComm. on the Judiciary, 104th Cong., 38 (1996); see also Dmitrieva, supra note 95, at245.127. See NII Copyright Protection Act of 1995: Hearings on H.R. 2441 Before theSubcomm. on Courts and Intellectual Property of the House Comm. on the Judiciary,104th Cong., 17 (1995).128. See id.129. See Dmitrieva, supra note 95, at 235.130. See NII Copyright Protection Act of 1995: Hearings on H.R. 2441 Before theSubcomm. on Courts and Intellectual Property of the House Comm. on the Judiciary,104th Cong., 35 (1995).131. See id. at 20.132. See WIPO Copyright Treaty and WIPO Performances and PhonogramsTreaty: Message from the President of the United States, Apr. 12, 1997, S. Treaty Doc.

Geneva during December, 1996.133 These treaties wereinternational in scope, and were intended to apply the regulatoryprovisions of the Berne Convention to the new digitalenvironment.134 While the WIPO treaties suggested thatcopyright liability should not apply to a person or entity serving asa conduit, who “provi[des] . . . physical facilities for enabling ormaking a communication,”135 the treaties did not specificallysuggest a standard of liability for ISPs, leaving this question forthe individual countries to decide.136Following the WIPO treaties, Congressional debate resumedto create a bill that would enact the treaties’ provisions andadequately address other copyright concerns, such as the properstandard of liability for ISPs. After several months ofCongressional debate, Congress passed the Digital MillenniumCopyright Act (DMCA),137 which incorporated an industryagreement between ISPs and copyright holders with an earlier billintended to codify the court’s decision in Netcom and overturnFrena.138 In enacting the Online Copyright InfringementLiability Limitation Act (OCILLA),139 Title II of the DMCA,Congress intended to balance the actual knowledge standardadvocated by ISPs with the strict liability standard advocated byrights holders by adopting an intermediate standard ofknowledge.140 This “intermediate” standard of knowledgeNo. 105-17 (1997).133. See id.134. See Mihály Ficsor, Towards a Global Solution: The Digital Agenda of theBerne Protocol and the New Instrument, in The Future of Copyright in a DigitalEnvironment 111-37 (P. Bernt Hugenholtz, ed., (1996)).135. See Agreed Statements Concerning the WIPO Copyright Treaty, WIPO Doc.No. CRNR/DC/96 (Adopted Dec. 20, 1996).136. See id.137. 17 U.S.C. §§ 1201-1332 (Supp. IV 1998).138. When Representative Coble introduced House Bill 3209, he included in hisintroductory remarks that he intended to codify the court’s decision in Netcom, so far asit established liability standards for ISPs, and wished to overturn the holding of Frena,“inasmuch as that case might apply to service providers, suggesting that such actscould constitute direct infringement.” See 144 Cong. Rec E160-01 (daily ed. Feb. 12,1998) (statement of Rep. Coble); see also Frena, 839 F. Supp. 1552 (M.D. Fla. 1993)(holding a bulletin board operator strictly liable for third party acts of infringementwhen he did not have actual or constructive knowledge of the posted infringement andremoved the material immediately once notified).139. 17 U.S.C. § 512 (Supp. IV 1998).140. See WIPO Copyright Treaties Implementation Act, and Online CopyrightLiability Limitation Act: Hearings on H.R. 2280 Before the House Subcomm. On Courts

imposed ISP liability both in instances where the provider hadactual knowledge of the third-party infringement, and whereconstructive knowledge could be inferred from an ISP’s awarenessof “facts or circumstances from which infringing activity isapparent.”141 This is more stringent than the pre-existingknowledge standard for contributory infringement, which imposedliability if the party knew or should have known that the materialwas infringing.142OCILLA establishes new parameters both defining andlimiting the liability that Internet service providers face for thirdpartyacts of copyright infringement on their services. Bydefinition, OCILLA applies to both traditional ISPs and nonprofitinstitutions of higher education in their capacity as Internetservice providers to students and faculty.143 The provisions ofOCILLA offer ISPs affirmative defenses whereby they can escapeliability for third-party acts of copyright infringement, whetherfacing direct, vicarious, or contributory liability.144 These safeharbor defenses are briefly described in the sections that follow.Passive ConduitDIRECT LIABILITYOCILLA offers a safe harbor for service providers whoperform only the function of a “conduit,” merely transmitting,routing, or providing connections for the digital communications ofothers.145 This provision recognizes the similarity between ISPsand “passive carriers” which were originally exempt under theCopyright Act, as they do not exercise control over the contents oftheir clientele.146 In order to take advantage of this safe harbor, aand Intellectual Property of the Comm. on the Judiciary, 105th Cong., 82 (1997).141. Id.142. The new knowledge standard has been defined as a “red flag” test. Accordingto the House Judiciary Committee, a red flag constitutes “information of any kind thata reasonable person would rely upon” in determining that material is infringing. SeeWIPO Copyright Treaties Implementation Act and Online Copyright InfringementLiability Limitation: Report on H.R. 2281 of the House Comm. on the Judiciary, H.R.Rep. No. 105-551, pt. 1, at 25 (1998). For a discussion of the new knowledge standardand how it should be applied with reverence to the appropriate legislative intent seeDmitrieva, supra note 95, at 251, 253-61.143. 17 U.S.C. § 512.144. See id.145. See 17 U.S.C. § 512(a)(1-5) (Supp. IV 1998).146. See 17 U.S.C.A. § 111(a)(3) (West Supp. 1999) (stating that activities of

provider must not initiate, select, or modify the content of thecommunication—meaning that providers must not determine themessage’s recipients or maintain a copy of the transmission on itsservice in such a manner that it can be accessed by others for alonger period than is reasonably necessary for the transmission totake place.147System CachingOCILLA offers a safe harbor to ISPs for the necessaryfunction of system caching, a technical procedure whereby the ISPmakes a temporary copy of information while transmitting itonline.148 To qualify for this defense, an ISP must meet thefollowing conditions: no modification of the material’s content;regular updating of the material in accordance with technicalstandards; no interference with the ability of technology to returncertain data to the original site; and no circumvention of passwordmechanisms.149 The ISP must also expeditiously remove thematerial alleged to be infringing upon notification from acopyright owner.150VICARIOUS OR CONTRIBUTORY LIABILITYInformation Residing on a System or Network at the Direction ofUsersOCILLA limits the liability of an ISP that stores third-partycontent that is completely under the control of that user.151 Thissafe harbor is contingent on the intermediate knowledge standard:the ISP must not have actual knowledge that the material oractivity is infringing, must not be aware of facts or circumstancesfrom which infringing activity is apparent, and must notfinancially benefit from the infringing activity.152 This defensealso requires the ISP to expeditiously remove or disable access topassive carriers should be limited to “providing wires, cables, or other communicationschannels for the use of others”).147. 17 U.S.C. § 512 (a)(1-5).148. See id. § 512 (b)(2).149. See id.150. See id. § 512 (b)(2)(E).151. See id. § 512 (c).152. 17 U.S.C. at § 512 (c)(1)(ii) and (d)(1)(B).

any infringing material either upon notification of a rights holderor upon gaining awareness of the prohibited activity.153Information Location ToolsOCILLA also offers a similar safe harbor for ISPs that provideinformation location tools, such as search engines, that link usersto other web sites indiscriminately.154 This defense is also limitedby the intermediate knowledge standard, and includes therequired notice and take down provisions mentioned above.155To be eligible for these liability limitations, ISPs are requiredunder OCILLA to designate an agent to receive notifications ofcopyright violations of their networks156 and must implement apolicy for the termination of subscribers’ accounts who repeatedlyact in violation of copyright law.157The safe harbor provisions limiting ISP copyright liability aredirectly related to specific functions performed by Internetproviders, and protect ISPs from liability only for the innocuoustransmission, copying, hosting, or linking of third-party contentmaterials that the provider does not itself create or control. Theseliability limitations are specific, and better assess the role of theInternet provider in infringing conduct than does the blanketimmunity for content provided by the CDA. The limitingprovisions listed in OCILLA reflect the careful scrutiny ofCongressional debate and lobbying on the part of both copyrightowners and Internet service providers.As a result of this lengthy process, the safe harbors outlinedin OCILLA have served as a model for other countries to follow incodifying copyright limitations on ISP liability. OCILLA’s limitingprovisions have not, however, been restricted to defining thelimitations on copyright liability in other countries. This paperwill now address the horizontal approach to ISP liability currentlybeing implemented in the European Union. This approach appliessafe harbors almost identical to those in OCILLA to all forms ofpotential ISP liability, creating a uniform, predictable minimum153. See id.154. See id. § 512 (d).155. See id. § 512(b)(2).156. See id. § 512 (c)(2).157. See id. § 512(I)(1)(A).

standard for each of the member states to follow in enacting theirown provisions.III ISP LIABILITY UNDER THE EU E-COMMERCE DIRECTIVEThe European Union E-Commerce Directive158 outlines a listof ISP safe harbors almost identical to OCILLA’s provisionslimiting copyright liability of Internet providers for third-partyacts of infringement in the United States.159 A significantdifference in the application of these two “statutory” provisions isthe comparative breadth of the safe harbor provisions. UnlikeOCILLA’s limiting provisions, which provide safe harbors forInternet providers only under copyright law,160 the EU Directivewill apply across the board to all areas of law involving ISPs.161This creates a unified, “horizontal” approach to determining theextent of Internet provider liability in Europe, and avoids theuncertainty of having different legal standards for determiningwhen ISPs face potential liability.162The E-Commerce Directive will significantly change the lawsof countries that have not independently developed statutoryprovisions limiting ISP liability163 or have adopted them, buthave, like the United States, made the provisions particular to onearea of law.164 By employing a horizontal approach to the issue of158. See Council Directive 00/31, 2000 O.J. (L 178).159. One aspect in which the E-Commerce Directive and OCILLA’s limitingprovisions differ is the Directive’s lack of a comparable safe harbor for ISPs functioningas information location tools. Article 24 of the Directive includes a re-examinationprocedure whereby the European Commission can re-evaluate the importance ofincluding a liability limitation for ISPs providing these linking services within threeyears of the Directive’s adoption. Some scholars have hypothesized that during thistime a separate safe harbor for information location tools will be added to the Directive.See Julia-Barcelo, supra note 11, at 115; Claus Köhler & Kai Burmeister, CopyrightLiability on the Internet Today in Europe (Germany, France, Italy and the E.U.), 1999Eur. Intell. Prop. Rev. 21(10) 485-99, 499.160. See Dmitrieva, supra note 95, at 240.161. See Julia-Barcelo, supra note 11, at 108.162. See id; see also Andre Lucas, Exploitation and Liability in the Light of MediaCoverage, 2001 Eur. Intell. Prop. Rev. 23(6), 275, 275 (discussing the horizontalapproach of the EU E-Commerce Directive and relating some of its advantages anddisadvantages).163. See Köhler & Burmeister, supra note 159, at 497-98 (discussing the lack ofexisting ISP liability limitations in France and Italy).164. See, e.g., Godfrey v. Demon Internet Ltd., 1999 WL 477647, (Q.B. Mar. 26,1999) (applying the U.K. Defamation Act of 1996). This Act imposes a variety oflimited liabilities for ISPs, provided that they lack knowledge of third party contentand exercise reasonable care under the circumstances. The law does not regulate

provider liability, the European Commission hopes to ensurepredictability for ISPs and promote harmonization between thevarious member states who must use the Directive as a minimumstandard for enacting their own legislation. As Julia-Barcelo165argues, this approach also eliminates the need for self-monitoring:A horizontal approach is appropriate . . . because theinformation that travels through on-line intermediaryfacilities such as cables, satellites and servers is just asequence of bits, a succession of zeros and ones which isnot linked to the real meaning of the information itself.Therefore, to apply different legal standards to differentmaterial flowing over their systems, on-lineintermediaries would actually have to convert all thosezeros and ones into text, music and images, and monitorand review each piece of such information. Imposingsuch an obligation on on-line intermediaries would bringthe Internet to a halt and would threaten the privacy ofinformation moving over it.166Julia-Barcelo’s comments highlight a valid concern withvertically applied ISP liability. For the American dual-standardliability scheme to properly enforce Congress’ stated intent inpassing both Acts,167 American Internet service providers must, insome form or fashion, monitor third-party content posted to theirsystems.168 It is this self-policing privilege that Congress soughtto protect by limiting provider liability for third-party content inthe CDA.169 Congress also advanced this goal by limiting providerliability for removing any third-party material that either theprovider or other users found offensive.170 While this liabilitylimitation does enable ISPs to freely monitor and police theirsystems, it does not require any such activity on their part.171 Ascopyright liability of ISPs. See id.165. See Julia-Barcelo, supra note 11.166. Id. at 108.167. See supra. text; see also 144 Cong. Rec E160-01 (daily ed. Feb. 12, 1998)(statement of Rep. Coble); Joint Explanatory Statement of the CongressionalConference Committee, H.R. Conf. Rep. No. 104-458 (1996), reprinted in 1996U.S.C.C.A.N. 124.168. See id.169. See id.170. See supra. text; see also 47 U.S.C. § 230(c)(2)(A).171. See Joint Explanatory Statement of the Congressional Conference Committee,H.R. Conf. Rep. No. 104-458 (1996), reprinted in 1996 U.S.C.C.A.N. 124.

a result, under the current interpretation of the CDA, a providergains the benefit of complete immunity without upholding anylevel of requisite responsibility, even when actual or constructiveknowledge of the problematic content is arguable.172While still not requiring monitoring,173 OCILLA’s limitingprovisions create liability for an Internet provider where the ISPmeets one of three requirements: actual knowledge of a user’sinfringement,174 awareness of facts or circumstances from whichthe infringement is apparent, or receipt of a financial benefit fromthe user’s infringing activity.175 If any of these provisions aremet, an ISP can still avoid liability under OCILLA by promptlyremoving the questionable material upon notice of the possibleinfringement.176 The EU E-Commerce Directive applies similarsafe harbor provisions horizontally to define the level of liabilityISPs face throughout all areas of law, including defamation.177Thus, the E-Commerce Directive avoids content monitoringconcerns not by granting wholesale immunity of the kind providedby the CDA, but rather by analyzing the facts of each case todetermine whether the ISP gains liability limitation by fitting intothe listed safe harbors. Where third-party content is concerned,this evaluation will focus on the amount of editorial control theISP exercised, the level of knowledge it possessed, and thefinancial benefit it gained. This method for determining ISPliability seems more in line with the requirement in commondefamation law of determining based on the facts of the casewhether or not a person or entity acts as a publisher or distributor172. See Zeran, 129 F.3d at 330-31.173. See DMCA Report on S. 2037 of the Senate Comm. on the Judiciary, Rept.105-190, 105th Cong. (1998).174. See text supra. See also supra notes 138-52 and accompanying text.175. See id.176. See id.177. One difference in the respective safe harbors provided by OCILLA and the E-Commerce Directive is the adoption of a “notice and take down” requirement as aprerequisite for obtaining safe harbor immunity. While OCILLA statutorily requiresthat potentially infringing material be promptly removed by an ISP after notification,the E-Commerce Directive includes no such per se requirement. See P. BrentHugenholtz, Caching and Copyright: The Right of Temporary Copying, 2000 Eur.Intell. Prop. Rev. 22(10) 482, 492. However, failure to expeditiously remove infringingmaterial after notification can still result in liability if a court finds, applyingunderlying general principles of the respective country’s law, that the ISP had actualnotice of infringement on their system and failed to take reasonable steps to remedythe situation. See id.; see, e.g., Godfrey, 1999 WL 477647 (discussed at length in textinfra).

of a defamatory statement.While safe harbor provisions employed by the EU E-Commerce Directive and OCILLA offer a more evaluative look atthe activities engaged in by Internet providers for determiningprovider liability in a defamation context, there remains oneaspect of OCILLA that conflicts with another fundamentalcharacteristic of U.S. defamation law. These conflicting provisionsare the “notice and take-down” provisions statutorily createdthrough OCILLA and the freedom of speech guarantee of the FirstAmendment.178IV. ISP LIABILITY UNDER OCILLA AND THE FIRST AMENDMENTTo retain OCILLA’s safe harbor provision limiting ISPliability for third-party information posted on its service, anInternet provider must promptly remove any allegedly infringingmaterial upon notification by a copyright owner.179 This featureof OCILLA arguably impedes the First Amendment right to freespeech, because the Act in effect requires an ISP to remove orblock access to information upon a mere allegation ofinfringement, without a required showing that an infringementhas indeed occurred. As Alfred Yen argued, OCILLA’s notice andtake-down provisions could have First Amendment implicationseven in a copyright context:. . . [T]he ambiguity of copyright law will make itimpossible to determine accurately whether manypotential infringements are, in fact, infringement . . .even if the subscriber’s behavior appears to be infringing,the ISP has no way of reliably knowing whether thesubscriber has permission for the use or the copyrightholder in question objects.These problems place ISPs in a quandary. If theymonitor their networks, they will undoubtedly stumbleacross potential infringements, but they also will beforced to decide whether to remove those potential178. While freedom of speech is a fundamental concern in Europe as well as in theUnited States, U.S. defamation law is subject to heightened scrutiny because it mustcomply with the First Amendment to the Constitution. The First Amendment reads, inpertinent part, “Congress shall make no law . . . abridging the freedom of speech . . .”U.S. Const. amend. I, cl. 2.179. See id.

infringements without secure knowledge of whether suchremoval is appropriate. Like risk averse newspaperpublishers, ISPs faced with this dilemma will likelyresolve any such doubts they have in favor of removingthe material in question from the Internet because suchaction minimizes claims of infringement. Such behavior,however, implies resolving doubts about the silencing ofspeech against speech.180While First Amendment concerns exist in a copyright context,they have been explored in greater detail in defamation law,where one of the central issues in creating libel laws has beenbalancing enforcement with the need to prevent any “chillingeffect” on constitutionally protected free speech.181 Arguably,imposing the notice and take-down requirement outlined inOCILLA in defamation cases now covered by the CDA would havean immediate chilling effect on free speech, because such aprovision would, in effect, require the immediate removal ofinformation as soon as an allegation as to its defamatory contentwas made to ensure that the ISP retained its safe harbor for thirdpartypostings.In Europe, the horizontal approach of the EU E-CommerceDirective addresses these notice and take-down concerns byomitting statutory notice and take-down provisions within the textof the Directive.182 However, it is arguable that the Directive’ssilent provisions nonetheless implicitly create a notice and takedownrequirement for ISPs to maintain their limited liabilitystatus. Because the Directive constitutes only minimumrequirements for different member states to adopt into their ownrespective intellectual property legal schemes, European countriesare still free to apply general legal standards to conclude that,once notice of possible infringement is perfected, any failure of anISP to expeditiously remove the material from the Internet resultsin a disqualification of the ISP’s safe harbor protection.183 Due tothe horizontal application of the Directive, this possible notice andtake-down result would apply across the board in all casesinvolving Internet service providers, including both copyright and180. See Yen, supra note 24, at 1871.181. See New York Times v. Sullivan, 376 U.S. 254, 279-81 (1964).182. See Hugenholtz, supra note 177, at 492.183. See id. at 492.

defamation.184 One case that illustrates the impact of a possiblenotice and take-down regime on ISP liability is the recent U.K.case, Godfrey v. Demon Internet Limited.185In Godfrey, a British math, physics, and computer sciencelecturer sued Demon, a prominent Internet service provider inEngland and Wales, over a defamatory newsgroup posting thatthe provider carried on its service.186 While the actual third-partyuser who originally posted the defamatory statement remainsunknown, it was established that the posting originated in theUnited States with a different ISP.187 Demon’s sole connectionwith the statement was the fact that it carried the particularnewsgroup containing the defamatory content on its service.188The statement in question was “squalid, obscene and defamatoryof the plaintiff” and purported to come from the plaintiff, althoughhis name was misspelled.189 When Godfrey discovered thedefamatory newsgroup posting, he notified the managing directorof Demon that the posting was a forgery and requested that it beremoved from Demon’s Use-net news server.190 Although Demonadmitted receiving the message, and could have “obliterated theposting,” they did not remove the allegedly defamatory materialfrom their server until it naturally expired.191In deciding Godfrey, the court applied the U.K. DefamationAct of 1996,192 which, like the DMCA, provides safe harbors or“defences” limiting ISP liability as a publisher of third-partycontent when the provider can show that he was not the “author,editor, or publisher” of the defamatory statement;193 that he took“reasonable care” in relation to the statement;194 and that he “didnot know, and had no reason to believe, that what he did causedor contributed to the publication of a defamatory statement.”195While the court concluded that Demon was not the actual184. See Julia-Barcelo, supra note 11, at 108.185. 1999 WL 477647 (Q.B. Mar. 26, 1999).186. See id. at *2.187. See id.188. See id.189. See id.190. See id. at *3.191. See Godfrey, 1999 WL 477647 at *3.192. See id. (quoting from the U.K. Defamation Act of 1996).193. See id.194. See id.195. See id. at *3.

publisher of the statement in question,196 it held that the providerwas ineligible for the liability limitation provided by the act.197Since Demon was notified of the defamatory statement and didnothing, the court concluded that the ISP had failed to exercisethe reasonable care required by the Act.198 The court also heldthat Demon, after notification by Godfrey, was in possession ofsufficient information to make the ISP aware that its actions werecontributing to the publication of a defamatory statement.199Because the ISP did not take affirmative action to remove theallegedly defamatory material after Godfrey’s notification, it lostthe liability limitation that the Defamation Act would haveotherwise provided it.200 As a result, the court held Demon liablefor the third-party defamatory statements posted on its service.201Although Godfrey was not decided under the provisionsrecited in the E-Commerce Directive, the Godfrey case is anexample of how strict adherence to a notice and take-downrequirement can result in chilling free speech. As Godfreyillustrates, an ISP under this regime must remove anyproblematic material as soon as it receives notice that theinformation is either defamatory or infringing copyright law toprotect itself from liability. Further, both the Directive andOCILLA recognize liability for ISPs when the provider shouldhave been aware of facts or circumstances making infringementapparent.202 If the reasoning applied in Godfrey is followed, ISPscould be held liable under these provisions for failing to remove196. See id.197. See Godfrey, 1999 WL 477647 at *3.198. See id.199. See id.200. See id. at *7.201. See id. In deciding Godfrey, the court cited the American defamation casesdealing with ISP liability before the CDA, and then distinguished the court’sapplication of the CDA’s complete content immunity for ISPs in Zeran, noting that theBritish Defamation Act “did not adopt this approach or have this purpose.” SeeGodfrey, 1999 WL 477647 at *8. The Godfrey court also distinguished Lunney’scommon law approach to limiting ISP liability by noting that under English commonlaw Prodigy would have been regarded as the publisher of its user’s third party contentrather than a passive conduit. See id. By this discussion, the Godfrey court illustratesthe massive liability that British ISPs faced under common law defamation principles,and how this liability has been significantly softened by recent statutory provisions likethe Defamation Act, which work to create safe harbors for ISPs.202. See 17 U.S.C. § 512 (c), (d) (Supp. IV 1998); Council Directive 2000/31/EC,2000 O.J. (L 178); see also Julia-Barcelo, supra note 11, at 110; Köhler & Burmeister,supra note 159, at 499.

material that appears questionable as regards copyright ordefamation law, without any requirement that ISPs affirmativelyestablish that such information is in fact unprotected. Such broadremoval discretion, with or without notice of alleged impropercontent or infringement, would impede the right of free speech forindividual Internet users.To rectify this problem while still applying the functionspecificprovisions of OCILLA and the E-Commerce Directive toAmerican defamation law, the notice-and-take-down requirementsoutlined in OCILLA could be softened in the defamation context.One possibility for softening OCILLA’s take-down requirementwould be requiring those claiming to be the victim of defamationto approach an ISP not with an allegation of inappropriate contenton its system, but rather with a court injunction, specifying thatthe speech is unprotected and should be removed from theInternet.203 Since the alleged victim of defamation carries theburden of proving that the other party’s speech is in fact false anddefamatory in a courtroom setting, such a showing should also beappropriate to restrict the open forum for free speech that existson the Internet. Requiring such a heightened showing to sanctionthe removal of material based on its content is consistent with theheightened freedom of speech that is guaranteed by the FirstAmendment.204 Also, since the heightened removal standardwould apply only where content was concerned, it would notimpede the privileges of copyright owners under the BerneConvention, which establishes that copyright owners shall not berequired to actively enforce their rights of ownership to retain itsprivileges.205203. Several scholars discussing the reality of a horizontal application of ISP safeharbors have endorsed the need for a court to weigh the competing interests of thevictims of defamation against those of Internet operators before resorting to automaticremoval and the resulting imposition on freedom of expression. See Lucas, supra note162, at 278; see also Dun & Bradsreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749,756-61 (1985) (discussing, among other things, the need to balance the compelling stateinterest in enforcing defamation law against the First Amendment’s interest inallowing the free flow of speech in public debate, and how this balance is altered by thestatus of the victim as a public or private figure and the subject matter of the speech);Gertz v. Robert Welch, Inc., 418 U.S. 323, 345-48 (1974).204. See U.S. Const. amend I.205. See WIPO Copyright Treaties Implementation Act, and Online CopyrightLiability Limitation Act: Hearings on H.R. 2281 and 2280 Before the House Subcomm.On Courts and Intellectual Property of the Comm. on the Judiciary, 105th Cong. 58(1997).

CONCLUSIONThe Internet has continually expanded over the past decade,becoming an invaluable medium for communication on a globallevel.206 To efficiently enforce defamation and copyright lawonline, the liability of both individuals and Internet providersmust be easily determined and to some degree predictable. Itshould also be reflective of the roles that individual users andInternet service providers actually play in the digital world.Sometimes ISPs exercise no control over material that they sendvia the Internet, and act only as a conduit. In other instances, therole of the ISP is much more involved, and its knowledge of thecontent it produces is arguably greater. The blanket immunitythat the CDA offers Internet providers allows ISPs to act aspublishers and distributors of online content without facingliability for this role. This provision is outmoded. Defamation lawcan be better enforced on the Internet by employing a unified setof specific safe harbor provisions for ISPs that relate to thedifferent aspects of their online activities. Such a unified,horizontal approach to ISP liability is being constructed effectivelyin Europe.207 OCILLA’s negotiated provisions limiting ISPliability in the copyright context accurately address the role of anISP in transmitting information, and can be adequately modifiedfor application in a defamation context without improperlychilling free speech online. Such a unified standard will bettermeet the needs of users, providers, and defamation law in theUnited States.Lucy H. Holmes206. See Jane C. Ginsburg, Putting Cars on the “Information Superhighway”:Authors, Exploiters, and Copyright in Cyberspace, 95 Colum. L. Rev. 1466, 1466 (1995).207. See Julia-Barcelo, supra note 11.

FBI’s Carnivore: Is the GovernmentEating Away Our Right of Privacy?INTRODUCTIONWith the birth of the Internet has come a virtual reformationof how human society communicates. The exact number ofInternet users is nearly incalculable, but recent assessmentsestimate that nearly 300 million people worldwide are currentlyonline.1 These users can travel relatively freely among themillions of currently active Internet sites.2 The growth of thiscommunications medium in the last decade has been tremendous,and promises to continue at such a pace. The United StatesDepartment of Commerce has reported that less than 40 millionpeople worldwide had access to the Internet in 1996.3 Thisnumber jumped to more than 100 million people by the end of1997.4 Further research by the Department has indicated thatthe number of people and businesses using the Internet doublesevery 100 days.5 Thus, the Internet presents unprecedentedopportunities for global communications and commerce. However,it also poses dramatic risks to personal privacy. Every dayAmericans use the Internet to access and transfer vast amounts ofprivate data. From electronic mail and business transactions toshopping habits, web surfing profiles can reveal detailedblueprints of how people live.6 As more of our lives are conducted1. See Gavin Skok, Establishing a Legitimate Expectation of Privacy inClickstream Data, 6 Mich. Telecomm. Tech. L. Rev. 61, 63 (1999).2. See id.3. See Thomas T. Reith III, Consumer Confidence: The Key To Successful E-Commerce In The Global Marketplace, 24 Suffolk Transnat’l L. Rev. 467, 486 n.3(2001).4. See id.5. See id.6. See Carnivore’s Challenge to Privacy and Security Online: Hearings onCarnivore Before the House Judiciary Subcomm. on the Constitution, 106th Cong.(2000) [hereinafter Carnivore’s Challenge] (statement of Alan B. Davidson, Staff237

online and more personal information is transmitted and storedelectronically, the result has been a massive increase in theamount of sensitive data available to interested third parties,including the government.7 The question thus presented iswhether our government has taken advantage of this availability.Although not conclusive, some evidence suggests that theutilization of electronic surveillance for monitoring criminalactivity has practically exploded over the past decade, arguablyreplacing more traditional investigatory tools.8 Current dataindicates that this pace is unlikely to diminish. The FBI aloneestimates that over the next decade, given planned improvementsin the digital collection and analysis of communications, itsrequests for electronic surveillance orders will increase 300percent.9 It appears that these planned improvements, as well asadditional surveillance requests, may well be underway.During Congressional hearings in April, 2000, Robert Corn-Revere, an Internet and communications lawyer in Washington,D.C., first divulged evidence of the existence of “Carnivore,” anelectronic surveillance device developed by the FBI.10 Respondingto concerns from various privacy groups and others, formerAttorney General Janet Reno ordered that an independenttechnical review of the system be completed by December 8, 2000,in order to substantiate Carnivore’s compliance with bothconstitutional provisions and federal statutory wiretap laws.11The review, completed by the IIT Research Institute (IITRI), adivision of the Illinois Institute of Technology, essentially foundthe system to be legally compliant and determined most concernsover invasion of individual privacy were unsupported.12Despite winning a favorable review by the university, theCarnivore device continues to raise strong concerns about privacyCounsel, Center for Democracy and Technology).7. See id.8. See id.9. See id.10. See John Hall, Privacy Beginning To Be A Top Issue, Rich. Times Dispatch,July 16, 2000, at F2.11. See Frank James, U.S. Seeks University Experts to Review FBI’s E-MailProbes, Chi. Trib., Aug. 25, 2000, at N20.12. See Institute’s Report on Carnivore Causes Uproar Among Critics, 4 Telecomm.Indus. Litig.Rep. 12 (2000).

and the legal limits of government surveillance.13 Many computerexperts believe that the scope of the study was too narrow, andthat serious technical questions remain about the ability ofCarnivore to satisfy Constitutional and statutory thresholds foronline security and safety.14 This note explores some importantlegal issues raised by the operation of Carnivore, including a briefdiscussion of the adequacy and conclusiveness of the IITRIanalysis as it pertains to each issue.Part I of this comment provides a basic assessment of theoperational aspects of the Carnivore system. Part II explores thefundamental legality of the Carnivore device from aConstitutional perspective. Part III examines Carnivore’s likelycompliance with current federal statutory wiretap laws as theyhave been interpreted and applied by the judiciary. Finally, theconclusion offers comments and proposals for how to strike abalance between privacy interests and government objectives.I. OPERATIONAL OVERVIEW OF CARNIVOREAs an initial matter, to understand the legal implications ofutilizing a surveillance system like Carnivore, one mustunderstand its basic operational capabilities. Although a fullydetailed description of how Carnivore works has not been madeavailable to the public, the general premise of the system is asfollows:Carnivore is a Windows NT based software program thatoperates upon connection to a network access point provided by aparticipating Internet Service Provider (ISP).15 To install thedevice, the FBI must present a valid court order to interceptelectronic communications of a target suspect.16 The court ordermay authorize capture of an entire communication, or it can belimited to addressing or routing information.17The FBI and the ISP install Carnivore at a point on the ISP’s13. John Schwartz, Computer Security Experts Question Internet Wiretaps, N. Y.Times, Dec. 5, 2000, at A16.14. See id.15. See The Carnivore Controversy: Hearings on Electronic Surveillance andPrivacy in the Digital Age Before the Senate Comm on the Judiciary, 106th Cong. (2000)[hereinafter Hearings 1] (statement of Sen. Patrick Leahy).16. See Big Sister – Janet Reno; An Orwellian Carnivore, Cincinnati Enquirer,Sept. 13, 2000, at A10.17. See Hearings 1, supra note 15 (statement of Sen. Patrick Leahy).

network where data from a suspect named in the court order islocated.18 As a technical matter, the FBI has conceded that theISP cannot provide an access point that can limit the Internettraffic flowing through the Carnivore device to only that of thenamed suspect.19 Because the Internet operates by breakingelectronic transmissions down into “packets” of data that arereassembled at a destination point, Carnivore must necessarilyseparate or filter the target suspect’s electronic transmissionsfrom other Internet traffic as it flows through the device.20Once the named suspect’s data is separated from other traffic,Carnivore routes the information to a second filter.21 As the datatravels through the second filter, the system makes a copy of all ofthe information and sends the original data to its desireddestination.22 On the basis of authorized search parameters,Carnivore segregates relevant and irrelevant data.23 Copied datathat is not relevant to the FBI investigation is purged.24 Trafficthat is relevant to the investigation and defined under the courtorder is sent to an archive system for permanent storage at anFBI facility.25With these functional characteristics in mind, thefundamental legality of the Carnivore device can be examinedfrom both a constitutional and federal statutory perspective.Because this comment focuses only on the inherent legitimacy ofthe system, assuming proper use, these analyses will be madeignoring any possibilities for misuse of the system on the part ofthe FBI.18. See id.19. See James X. Dempsey, Does Carnivore Go Too Far?, Network World, Oct. 30,2000, at 73; see also Dan Eggen & David A. Vise, More Questions Surface about FBISoftware; Wiretap Program Can Archive All Internet Communications, Nov. 18, 2000,at AO3.20. See Hearings 1, supra note 15.21. Id.22. See The Carnivore Controversy: Hearings on Electronic Surveillance andPrivacy in the Digital Age Before the Senate Comm. on the Judiciary, 106th Cong.(2000) [hereinafter Hearings 2] (statement of Donald Kerr, Asst. Director, FederalBureau of Investigation).23. Id.24. Id.25. See Hearings 1, supra note 15.

II. CONSTITUTIONAL ANALYSISA. Constitutional Issues PresentedThe Fourth Amendment to the United States Constitutionprovides that:[T]he right of the people to be secure in their persons,houses, papers, and effects, against unreasonablesearches and seizures, shall not be violated, and nowarrants shall issue, but upon probable cause, supportedby oath or affirmation, and particularly describing theplace to be searched, and the persons or things to beseized.26The traditional judicial interpretation of a FourthAmendment search of persons, houses, papers and effects requiredsome type of physical invasion on the part of the government.27However, the scope and interpretation of Fourth Amendmentapplication was broadened in 1967 when the Supreme Court of theUnited States determined that a search, as defined by theAmendment, could be accomplished via government orchestratedelectronic surveillance.28 Thus, the constitutional implicationscreated by the utilization of the Carnivore surveillance devicemust be analyzed and reviewed in such context. The primaryFourth Amendment concerns raised by Carnivore can be narrowedto two issues.Issue 1 – Warrantless Search of Persons Not Named In CourtOrderBecause the FBI has conceded Carnivore’s inability to isolatea target suspect’s transmissions as they flow through its firstfiltering point,29 is the device able to conduct an unauthorized,warrantless search of communications of persons not subject tothe authorized surveillance?26. U.S. Const. amend. IV (emphasis added).27. See Olmstead v. United States, 277 U.S 438 (1928); see also Clifford S.Fishman, Electronic Tracking Devices and the Fourth Amendment: Knotts, Karo andthe Questions Still Unanswered, 34 Cath. U.L. Rev. 277, 286 (1985).28. See Fishman, supra note 27, at 395.29. See Dempsey, supra note 19, at AO3.

Issue 2 – General Warrant Search of Target SuspectBecause the FBI has stated Carnivore copies and reviews, insome manner, all electronic transmissions attributable to asuspect named in a court order,30 does the device conduct ageneral search of the suspect’s electronic communications notspecified in the order?1. Examination of Fourth Amendment ConcernsThe FBI has established that the Carnivore device readscertain “addressing” and “transactional” information fortransmissions flowing through Carnivore’s first filter point.31Much of this information is not authorized in a valid court orderand has no connection to the approved purpose of the electronicsurveillance.32 The IITRI report seems to confirm that whenCarnivore is used for trap-and-trace surveillance to interceptsimply the “to” and “from” information on an e-mail, the softwaregives investigators more information than may be permitted bythe court order, including the length of e-mail messages.33Carnivore intercepts such material in one of two ways. First,the device may receive the data without examining the actualbody of the transmission.34 Second, the system may view theaddressing information only in conjunction with the body of thecommunication.35As an initial matter, an individual raising a FourthAmendment challenge to a government search must show that theactions of the government infringed upon his legitimateexpectation of privacy.36 If the individual does not establish thisreasonable expectation, no Fourth Amendment search, andtherefore violation, has occurred. The judicially created test fordetermining whether an individual has a reasonable expectationof privacy in an electronic communication is a two-prong30. Id.31. Hearings 2, supra note 22.32. Id.33. Jon Baumgarten, Official Report on Carnivore Supports E-Mail Surveillance,5 Cyberspace Law. 23 (2001).34. Carnivore’s Challenge, supra note 6.35. Id.36. See Skok, supra note 1, at 71.

application developed in Katz v. United States.37 The first prongof the analysis inquires as to whether the individual has asubjective expectation of privacy in the communication. Thesecond prong focuses on whether society will recognize suchexpectation.38 Thus, in order to establish that Carnivore isconducting an invalid warrantless search of information notsubject to a court order, both prongs must be satisfied.Subjective ExpectationGenerally, to possess a subjective expectation of privacy, anindividual merely has to exhibit that he has some actual desire tokeep the communication secluded or away from public access.39With respect to most Internet transmissions, electronic mail inparticular, one can assume that most people have a strong actualexpectation of privacy in the content of messages they send andreceive. Safeguards such as login names, passwords andencryption procedures lend credibility to this assumption.However, the existence of an actual expectation may becomeless convincing when the communication is what has traditionallybeen regarded as “addressing” or “transactional” information. Theability of an individual to successfully claim an actual expectationof privacy in electronic addressing and transactional data hasbeen addressed by the Supreme Court on two definitive occasions.In Smith v. Maryland,40 the Supreme Court held that a person “inall probability” does not have an actual expectation of privacy inthe telephone digits he dials because he voluntarily conveys suchinformation to a third party, the telephone company.41 Thus, theindividual assumes the risk that such information will bereleased. Similarly, in United States v. Miller,42 the Court heldthat an individual has no subjective expectation of privacy in bankrecords because he freely conveys this information to his bank inthe ordinary course of business.43 The real issue to be addressedhere is whether such reasoning is applicable in the Internet arena.37. 389 U.S. 347 (1967).38. See id.39. See Smith v. Maryland, 442 U.S. 735 (1979).40. 442 U.S. 735 (1979).41. See id. at 742.42. 425 U.S. 435 (1976).43. Id. at 442.

Internet addressing and transactional material is generallyeither Uniform Resource Locator (URL) data or Internet Protocol(IP) addresses.44 A URL is the “electronic address” a person typeswhen sending an electronic mail message.45 “JohnDoe@aol.com”is an example of a URL address. An IP address is a computer’spersonal identification number that accompanies any electronictransmission that is sent over the Internet from that particularcomputer.46 An example of an IP address is 207.226.3.43.As the Internet has become more and more of a pervasive partof everyday life, some discussion has suggested that a traditionalanalysis, such as that invoked in Smith and Miller, should not beapplied to the Internet forum. The basis for such a proposition isthe theory that Internet addressing information, unlike moretraditional addressing or transactional material, is much morerevealing of content. For instance, if an individual logs onto aparticular Website and requests information, literature, etc., therequest message will be sent to that particular Website’s URLaddress, or mailbox.47 Therefore, if a person logs onto“WforPresident.com” and requests information about thepresidential campaign, how to donate, etc., this request will besent to the “WforPresident” Website URL. Unlike telephonedigits, which when viewed alone and in the absence of furtherinvestigation reveal little if any element of content, a simpleInternet message such as this could expressly illustrate a person’spolitical affiliations. In fact, unlike telephone numbers, theinterception of URL information can give law enforcement a fairlycomprehensive picture of the individual’s interests and activitiesonline.48 The FBI could know what type of books an individualreads, his romantic and artistic interests, and much more.49Advocates against the application of traditional reasoning arguethat such information is more analogous to a telephoneconversation than the digits dialed and thus should receive Fourth44. See Carnivore’s Challenge, supra note 6.45. See Tim Wyatt, Secure Shopping, Dallas Morning News, Apr. 27, 2000, at 3J;see also J. Timothy Hunt, Moving Target, Nat’l Post, Oct. 1, 2000, at 48.46. Carnivore’s Challenge, supra note 6.47. Id.48. See James X. Dempsey, Communications Privacy in the Digital Age:Revitalizing the Federal Wiretap Laws to Enhance Privacy, 8 Alb. L.J. Sci. & Tech. 65(1997).49. See id.

Amendment protections.50Substantial evidence of this protective approach makes itplausible that Internet users have an actual expectation of privacyin URL information being intercepted by Carnivore. In fact, manyprivacy groups are concerned about the potential of Carnivore toreceive and record such information.51 For example, ElectronicPrivacy Information Center (EPIC) filed a Freedom of Informationlawsuit against the FBI in July 2000, requesting release ofCarnivore’s capacity to view such information.52 The organizationwas successful in obtaining over 2,000 Carnivore relateddocuments from the Justice Department.53 However, finding theprovided information inconclusive, EPIC filed a motion in theDistrict of Columbia in August, 2001 requesting the deposition ofseveral FBI officials in order to acquire further informationregarding the device’s ability to retrieve such data.54 Also, privatesoftware companies, in response at least in part to Carnivore’sability to intercept URL information, have developed URLblocking software to prevent such access.55 Similarly, many suchcompanies have developed, or are in the process of developing lessinvasive surveillance devices, emphasizing a less intrusiveprocedure and enhancing privacy protections.56Alternatively, it would seem that a strong argument couldreasonably be made in favor of a traditional application of theCourt’s reasoning in Smith and Miller with respect to Internetaddressing information. Although, very few courts haveaddressed the applicability of such reasoning to the Internetforum, the courts seem to agree that a traditional approach isappropriate. Proponents of this position state that, like telephone50. See Randolph S. Sergent, A Fourth Amendment Model for Computer Networksand Data Privacy, 81 Va. L. Rev. 1181, 1201 (1995) (discussing how individuals do notassume the risk that the content of their telephone conversations will be disclosed by athird party (e.g. telephone company)).51. See Michael J. Miller, The Ever-Expanding Browser, PC Magazine, Oct. 3,2000, at 7.52. See Maria Mosquera, Privacy Group Wants Speedier Carnivore Disclosure,Techweb News, Aug. 18, 2000.53. See Epic Wants A Closer Look At ‘Carnivore’, Nat’l J. Tech. Daily, Aug. 13,2001.54. See id.55. See Accelerated Networks Introduces Secure Multiservice Broadband Access;DSLcon, Bus. Wire, Sept. 19, 2000.56. See Ann Harrison, Don’t Like Carnivore? How About Altivore? Open SourceCode Version of E Mail Sniffer in the Works, Computerworld, Sept. 25, 2000, at 12.

digits and bank records, URL information is voluntarily submittedto third parties. Internet users must realize that they “convey”their URL information since this information must travel throughthe ISP’s network to reach whatever destination is desired.57Also, ISPs require Internet users to choose a unique URL addressin order to facilitate delivery and transmission of electroniccommunications, much like a telephone company does when itassigns telephone numbers.58 The analogy can go further. InSmith, the Court determined that the individual voluntarily turnsover addressing information to the telephone company and thusassumes the risk that the company will reveal the information tolaw enforcement. This assumption must be made even thoughtelephone companies normally do not record every number dialedby a customer.59 Thus, although it is quite probable that an ISPwould not record every transmission made, every transmissionsent by a user must flow through the ISP’s server, and thereforethe Internet user must assume the risk that such information willbe revealed to law enforcement.60Applying the reasoning of Smith and Miller to IPs, addressinformation may be an easier position to argue for thosesupporting the traditional approach. Unlike URL information, IPaddresses are numerical, therefore much more like telephonenumbers and contain little, if any, content element. However, aswith URL data, some effort has been made to protect suchinformation from public access. On October 12, 2000, staterepresentative Gene Green introduced a bill that would prohibitISPs from placing software “cookies,” a type of tracking andidentification device, on consumer IP addresses.61 Also, theAmerican Bar Association has created a new website to helpconsumers avoid cookies as they shop online.62This practical evidence suggests that individuals57. Smith, 442 U.S. at 742 (reasoning that an individual realizes that he mustconvey phone numbers he dials to his telephone company since his call is completedthrough the company’s switching equipment).58. See Paul Schwartz, Privacy and Democracy in Cyberspace, 52 Vand. L. Rev.1609, 1610-11 (1999).59. Smith, 442 U.S. at 735.60. See Skok, supra note 1, at 67.61. See Drew Clark, Privacy: Green Introduces Online Privacy Policy Measure,Nat’l J. Tech. Daily, Oct. 12, 2000.62. See Sara Hazlewood, ABA Offers Tips to Help Consumers Protect Themselveson the Internet, 17 Bus. J. 7 (Dec. 3, 1999).

communicating over the Internet possess some level of asubjective expectation of privacy in the URL and IP informationthat Carnivore uses to conduct electronic surveillance. However,the real question in the analysis is whether society is ready torecognize such an expectation.Society’s ExpectationThis second, and more difficult prong of the two part Katz testto satisfy, addresses whether society is prepared to recognize areasonable expectation of privacy in the electronic communication.The Supreme Court has stated that society is not willing torecognize an individual’s expectation of privacy unless thatexpectation is “objectively reasonable.”63The Katz court defined objective reasonableness in the form ofa two-factor application. The first factor addresses whether theindividual attempted to deny public access to hiscommunication.64 The second factor focuses on whether lawenforcement has intercepted the content of the individual’stransmission.65Katz involved the warrantless recording of a telephoneconversation an individual had in a public phone booth that hadbeen bugged by law enforcement officials. Applying these twofactors, the Court determined that the individual subjected to thesurveillance had an objectively reasonable expectation of privacyin using the phone booth because he had taken steps to denypublic access to his conversation.66 Also, law enforcement had fullaccess to the content of the individual’s phone conversationbecause they had placed a recording device on the exterior of thebooth itself.In the years since Katz, at least with respect to those caseswhere the Court has been presented with more traditionalparameters,67 the Supreme Court’s reasoning and application of63. California v. Greenwood, 486 U.S. 35 (1988).64. Katz, 389 U.S. at 351.65. Id. at 354.66. Id. at 347.67. See Allegra Knopf, Privacy and the Internet: Welcome to the Orwellian World,11 J. Law & Pub. Pol’y 79, 83 (1999) (discussing how the Katz framework has appliednicely to situations where society holds conventional notions of public and privateplaces).

determinative factors has been further verified.68 Where there islittle, if any established parameters, such as the FourthAmendment’s application regarding the Internet, the analysis hastended toward a narrow application of the Katz reasoning.69 Infact, the few judicial decisions expressly addressing Internettransmissions have applied the Katz reasoning very narrowly.70For instance, in United States v. Charbonneau,71 the court heldthat an individual’s expectation of privacy in electronic mail wassignificantly less than that held in posted mail.72 Also, in UnitedStates v. Kennedy,73 the court established that society would notrecognize an expectation of privacy in information a user passesonline to an ISP when contracting for Internet service.74Thus, the judicial decisions addressing personal privacy onthe Internet strongly suggest that a narrow interpretation of theKatz analysis will be applied to determine if society will recognizean expectation of privacy in URL and IP information. The firstinquiry would undoubtedly address whether Carnivore enablesaccess to the content of the electronic communications of personsnot named in a court order. It is very improbable that, under anarrow Katz application, a court would determine URL and IPinformation to be “content” because each serves a function similarto that of a telephone number. Also, if the judiciary is not willingto recognize an equal expectation of privacy in electronic mail tothat of posted mail, the probability of successfully arguing that theprotection of information can be intercepted by law enforcementwithout intrusion into the body of the electronic communication itaccompanies is minimal.The second inquiry in the Katz analysis would certainly focuson whether the individual attempted to restrict third party accessto his URL or IP information. Again, the judiciary will likelyconclude that users of the Internet must realize that they “convey”their URL information since this information must travel throughthe ISP’s network to reach whatever destination is desired.75 The68. See Minnesota v. Carter, 525 U.S. 83 (1998); see also O’Connor v. Ortega, 480U.S. 709 (1987).69. See Knopf, supra note 65, at 83.70. See Skok, supra note 1, at 72.71. 979 F. Supp. 1177 (S.D. Ohio 1997).72. Id. at 1184.73. 81 F. Supp. 2d 1103 (D. Kan. 2000).74. Id. at 1110.75. See Smith, 442 U.S. at 742 (reasoning that an individual realizes that he must

court will likely, as it has often done when addressing FourthAmendment applicability to technological advances, attempt toanalogize the situation with a more traditional setting. In thiscase, that traditional setting will undoubtedly be that establishedin Smith and Miller.From this examination, it seems unlikely that, even if anindividual can establish that he has a subjective expectation ofprivacy in URL and IP information attached to his electronictransmissions, the judiciary will determine that society isprepared to recognize such expectation. Thus, Carnivore’sinterception of URL and IP information relating to individuals notnamed in a court order is not likely a search as determined underFourth Amendment principles.If, in the alternative, the judiciary were to find the existenceof both a subjective and societal reasonable expectation of privacyin URL and IP data, Carnivore would undoubtedly be conductinga search of such information, and the next step in theconstitutional analysis would be to determine whether the searchwas a valid one.2. The Viewing of Addressing Information Inside Body ofTransmissiona. Is the Governmental Action a Fourth Amendment Search?Electronic addressing information, such as URL or IP data,can generally be intercepted by Carnivore without entering thebody of the accompanying electronic transmission.76 However,there may be certain instances in which such information islocated only within the body of the communication, such as the“TO:” line within an electronic message. This situation may occurwhen the particular software used to transmit the communicationdoes not reveal the URL address to the ISP, but merely transmitsthe IP address, or computer identification number.77 Forexample, if the FBI has a court order to search for a particularURL, the device may have to locate this information on the “TO:”line of the transmission if the only information sent from theconvey phone numbers he dials to his telephone company since his call is completedthrough the company’s switching equipment).76. See Hearings 2, supra note 22.77. See Carnivore’s Challenge, supra note 6.

computer to the ISP is its numerical IP address. Thus, in thesecases it is highly probable that Carnivore will enter the body oftransmissions sent or received by people not subject to a courtorder as they flow through the first filter point.Some argue that this invasion, however slight, is anunauthorized warrantless search. Those who support thisposition rely on the Supreme Court’s reasoning in Arizona v.Hicks.78 In Hicks, the Court held that a warrantless searchpremised on no judicially created exception will be deemedviolative of the Fourth Amendment even if the governmentalintrusion is minimal. The Court based its decision in part on thefact that such intrusions may give law enforcement access toinformation that they would not otherwise be legally entitled.Applying the Hicks reasoning, when Carnivore is searching forURL information in the body of an electronic transmission,although the intrusion is slight, the FBI may have access toinformation within the communication that it would not otherwisebe legally entitled.Although this reasoning has not yet been applied by thecourts to Internet communications, there is some evidence thatsuggests that the judiciary would not favor this approach. In1997, the Communications Assistance for Law Enforcement Act of1994 (CALEA)79 was amended to allow telecommunicationscompanies, in certain instances, to provide the full content ofcustomer communications to the government. This is true evenwhen the government is only authorized to intercept addressinginformation.80 The amendment relies on law enforcement tosegregate the addressing information from the content of themessages. No court has yet deemed the provision to be78. 480 U.S. 321 (1987) (concerning a situation where law enforcement officerswere legally on private premises investigating a shooting when an officer noticed andcertain stereo equipment within the home that he suspected was stolen. The officerslightly moved the equipment in order to obtain its serial number. The Supreme Courtheld that although law enforcement was legally on the premises, the movement of theequipment constituted a warrantless search because an action was taken by the officerto obtain information not otherwise available to him.); see also Carnivore’s Challenge,supra note 6 (statement of Alan B. Davidson, Staff Counsel, Center for Democracy andTechnology).79. 47 U.S.C. § 1001 (1994 & Supp. V 1999) (imposing a statutory obligation ontelecommunications providers to provide assistance to law enforcement when presentedwith valid legal authorization; although CALEA regulates telecommunications, it hasnot yet been amended to include Internet Service Providers).80. See Dempsey, supra note 46, at 97-98.

unconstitutional.81 Although CALEA is not applicable to ISPs,the amendment and its judicial acceptance is significant in that itillustrates again the trend of the judiciary to place a lower valueon, not only addressing information, but electroniccommunications in general. Therefore, based on the currentinclination of the courts, Carnivore’s invasion into the body ofcommunications not subject to investigation is not likely to beconsidered a search for Fourth Amendment purposes.b. Assuming the Government Action is a Search, is it a ValidSearch?Generally, a search is valid under the Fourth Amendment if itis conducted via a warrant based on probable cause, or if it isreasonable.82 In implementing Carnivore, the FBI obviously hasno warrant authorizing the interception of electroniccommunications of persons not named in a court order. Thus, theinquiry becomes whether the FBI’s intrusion on the individual’sexpectation of privacy is reasonable.83 In determining whether agovernmental search is reasonable, courts will generally firstattempt to determine whether the Framers of the FourthAmendment would regard such action as unlawful.84 Where thisinquiry yields no definitive answer, the judiciary will ordinarilyevaluate the search under standards of reasonableness that havebeen developed though judicial determinations in the area ofFourth Amendment jurisprudence.85Determining the reasonableness of the FBI’s action regardingCarnivore based on standards considered by the Framers mayprove to be extremely difficult, if not impossible. The concept ofsearches of electronic communications was not a consideration ofthe drafters of the amendment. Under standards that have beendelineated through numerous Fourth Amendment judicial81. See id.82. See Frank W. Miller, et al., The Police Function 260 (6th ed. 2000).83. See Skok, supra note 1, at 71.84. See Florida v. White, 526 U.S. 559, 562-63 (1999); Wilson v. Arkansas, 514U.S. 927, 931 (1995); California v. Hodari, 499 U.S. 621, 624 (1991); Carroll v. UnitedStates, 267 U.S. 132, 149 (1925).85. See Wyoming v. Houghton, 526 U.S. 295, 299-300 (1999); Vernonia Sch. Dist.47J v. Acton, 515 U.S. 646, 652-53 (1995); see also Carroll, 267 U.S. at 149 (“TheFourth Amendment is to be construed in light of what was deemed an unreasonablesearch and seizure when it was adopted, and in a manner which will conserve publicinterests as well as the interests and rights of individual citizens.”).

decisions, however, warrantless searches are generally recognizedas reasonable in a limited number of situations. Withoutenumerating all possible circumstances, the most plausibleexception to the warrant requirement likely to be presented by theFBI to justify the operation of Carnivore is the existence of exigentcircumstances.86 The judiciary has identified an exigentcircumstance to be one in which there is the danger that evidencewill be destroyed if law enforcement takes time to obtain awarrant.87 Also, an exigent circumstance exists if lawenforcement have a reasonable belief that the safety of an agent orother innocent individuals is an issue.88 An argument suggestingexigent circumstances in an electronic surveillance situation isweakened by the Supreme Court’s determination that suchsurveillance can be accomplished without prior notification to theindividual under investigation.89 Thus, the risk that a suspectsubject to surveillance by Carnivore will destroy, or in this case,fail to transmit, electronic communications in anticipation of FBIintervention is not likely to be viewed as “reasonable” under thecurrent interpretation of the Fourth Amendment.Further, as an alternative to searching for a suspect’s data byidentifying certain URL information, law enforcement couldobtain a warrant to conduct a search solely using IP information.This type of information always accompanies electronictransmissions and is always present in the header addresslocation. Thus, the need to enter the body of the transmission iseliminated. However, such an alternative may not be a practicalone. Most IP addresses are temporary, and are assigned to adevice when it is connected to an ISP’s network.90 These addressesfrequently change and thus likely cannot be electronicallymonitored for any significant period of time.91c. General Warrant Search of a Target SuspectThe second constitutional concern presented by the operationof Carnivore focuses on the interception and collection of electronic86. Miller, supra note 73, at 261.87. Id.88. Id.89. See Katz, 389 U.S. at 355 n.16; see also Ker v. California, 374 U.S. 23, 37-41(1963).90. See Latest IP Prompts Net Privacy Fears, Computing, Oct. 28, 1999, at 14.91. See id.

data attributable to the target suspect of the electronicsurveillance. Carnivore functions by copying all electronic datarelating to a target individual that is sent through a particularISP’s network. When intercepting and copying such data,Carnivore can employ an “Internet pen register function” or an“Internet wiretapping function.”92 The constitutional issuepresented is whether either of these functions present the risk ofcreating a general warrant search of the target individual’scommunications.Internet Pen Register FunctionCarnivore’s pen register function operates by copying andstoring only authorized source and destination data, otherwisetermed addressing information.93 This information, such as “to”and “from” electronic mail addressing material, is expresslydescribed in a court order, and the FBI must store only suchspecified information.94 All other copied data must be purged.95As with communications relating to persons not named in acourt order, there are certain situations where Carnivore mayenter the body of a transmission pertaining to a named suspect tofacilitate the search for URL information. Privacy groups andothers are concerned that these circumstances transform anotherwise valid court order into a general warrant because theFBI can now search practically all of the communications sent toor from the suspect.96 This is an important function when oneconsiders the extremely low standards for acquiring courtauthorization to intercept source and destination information forelectronic communications.97 The standard, analogous to that of asubpoena or telephone pen register, holds that a judge mustapprove practically any surveillance request that can reasonablybe expected to produce evidence relevant to an ongoing criminalinvestigation.98 However, to obtain authorization to intercept the92. See Carnivore’s Challenge, supra note 6.93. See Hearings 2, supra note 22.94. Id.95. Id.96. See Donna Howell, Security & Privacy Studies Set to Flesh Out FBI’sCarnivore, Investor’s Bus. Daily, Aug. 28, 2000, at 8.97. See 18 U.S.C. §§ 3122-23 (1994).98. See id.

content of certain electronic communications, the FBI must meetthe much higher standard applicable to wiretaps.99 It is thereforeplausible that Carnivore’s pen register function allows the systemto operate under the impermissive authority of a general warrant.Originally, Fourth Amendment protections were designedprimarily to protect individuals from the intrusive power wieldedby general warrants.100 However, modern judicial interpretationof the Amendment has seemed to dilute these principles.101Nowhere in Fourth Amendment jurisprudence is this moreapparent than in the judiciary’s lenient interpretation of theminimization requirements mandated in the federal wiretaplaws.102 Substantial evidence suggests that the interception ofunauthorized information attributable to a target suspect duringthe conduct of an electronic surveillance is rarely even frownedupon by the courts.103 In fact, law enforcement is often affordedjudicial tolerance when certain technical or practical difficultiesinhibit their ability to minimize the interception of such data.104Reasoning from this perspective, Carnivore’s inability toconsistently intercept addressing or routing information withoutentering the body of an electronic transmission is likely to beallowed by the courts and not viewed as an unconstitutionalgeneral warrant search of the target suspect’s communications.Internet Wiretap FunctionCarnivore’s wiretapping function operates by searching all ofthe target suspect’s Internet communications for key words orphrases that are described in a court order.105 Although the FBIcopies all transmissions pertaining to the suspect, it must storeonly communications containing such key words or phrases, andpurge all other information.106 This “stored” data is sent to an99. See 18 U.S.C. § 2510 (1994 & Supp. V 1999).100. See Christopher E. Torkelson, The Clipper Chip: How Key Escrow Threatens toUndermine the Fourth Amendment, 25 Seton Hall L. Rev. 1142, 1175 n.90 (1995).101. See id; see also Larry Downes, Electronic Communications and the Plain ViewException: More “Bad Physics,” 7 Harv. J. Law & Tech. 239, 278 (1994).102. See Dempsey, supra note 46, at 77.103. See Scott v. United States, 516 F.2d 751 (D.C. Cir. 1975); see also UnitedStates v. Ozar, 50 F3d. 1440 (8th Cir. 1995).104. See id.105. See Carnivore’s Challenge, supra note 6.106. See id.

FBI facility for subsequent examination by FBI agents.107Similar to the Internet pen register function, advocates ofgreater privacy protections on the Internet are worried that thisprocedure may also present a danger of a general warrant.108Those promoting this position believe there is a risk that atransmission could contain the specified “key word or phrase” butnot be relevant to the government’s investigation.109 For example,if a key word such as “drugs” is authorized, the FBI might collectand review a target suspect’s communication consisting ofanything from a request submitted to a website concerning newdrugs available for AIDS patients, to cures for the common cold.The IITRI report seems to concede this possibility. It states thatwhile the system is designed to, and can, perform fine-tunedsearches, it is also capable of broad sweeps.110 Incorrectlyconfigured, Carnivore can record any traffic it monitors.111 Thereport goes on to state that this possibility may be increased dueto the lack of adequate audit trail provisions or safeguards againstcommon configuration errors.112However, like the pen register function, the relaxedminimization requirements applied by the judiciary to moretraditional wiretap investigations are likely to be similarly appliedto Internet taps. Therefore, the FBI’s erroneous collection andstorage of target suspect material not relevant to the surveillanceis probably constitutionally permissible.Upon analysis, it appears that the FBI’s implementation andutilization of Carnivore is in compliance with Constitutionalprinciples, at least so far as these principles are currentlyinterpreted. The next stage of the legal analysis addresseswhether or not the application of Carnivore complies with federalstatutory wiretap laws.III. FEDERAL STATUTORY ANALYSISThe legitimacy of the Carnivore surveillance system can alsobe analyzed in a statutory context. The focal points of such an107. See Hearings 2, supra note 22.108. See Carnivore’s Controversy, supra note 6.109. See id.110. See Brian Krebs, Study Calls For Stronger Audit Trail In FBI’s Carnivore,Newsbytes, Nov. 21, 2000.111. See id.112. See id.

analysis are the privacy protections included in such provisionsthat add to the constitutional minimum. Once these addedsafeguards are determined, an assessment of how the judiciary islikely to interpret and apply such provisions must be made. Thefollowing presents a commensurate examination of the federalstatutory provisions applicable to Carnivore, as well as theprobable judicial opinion as to Carnivore’s compliance with suchrequirements.In 1968, one year after Katz, Congress passed Title III of theOmnibus Crime Control and Safe Streets Act.113 This actpermitted non-consensual wire or oral eavesdropping by thegovernment as long as the surveillance was done pursuant to therequirements delineated in the act.114 In 1986, Congressamended Title III by passing the Electronic CommunicationsPrivacy Act (EPCA).115The ECPA was intended to enhance Title III by establishingdefinitive rules for electronic surveillance.116 In drafting theECPA, Congress included certain privacy “protections” thatunderlie the essential purpose of the statute. These protectionscan be grouped into three categories.117 First, any attemptedelectronic surveillance is subject to stringent ex parte judicialreview. Second, those conducting the electronic surveillance mustminimize the interception of non-pertinent information during thesurveillance. Third, the conduct of and results derived from anauthorized electronic surveillance may be subject to a stringentadversarial review after the surveillance has been completed.118An analysis will show that these protections, althoughembodied in the language of the ECPA, have been significantlydiminished through judicial interpretation and action. Thus, theutilization of Carnivore, arguably an extremely invasiveinvestigative tool, is likely to be in full compliance with the federalstatutory requirements of the ECPA.Ex Parte Judicial Review113. 18 U.S.C. § 2510 (1994 & Supp. V 1999).114. See id.115. See Christopher Slobogin, Criminal Procedure: Regulation of PoliceInvestigation 190 (2d ed. 1998).116. 18 U.S.C. § 2518 (1986).117. See Dempsey, supra note 46, at 85.118. See id.

EPCA Section 2518 provides that upon an application for asurveillance order, the issuing judge may enter an ex parte orderauthorizing interception of the electronic communications.119 Inissuing such order, the judge must determine primarily that (1)there is probable cause for belief that an individual is committing,has committed, or is about to commit a specific offenseenumerated in the statute; (2) there is probable cause for beliefthat particular communications concerning that offense will beobtained through such interception; and (3) normal investigativeprocedures have been tried and have failed or reasonably appear tobe unlikely to succeed if tried or may be too dangerous.120The FBI has applied for and received approximately twentyfivecourt orders approving the use of Carnivore in the past twoyears.121 Although it is unclear what information was presentedto obtain such surveillance orders, the majority of courts havegenerally applied the same sufficiency of evidence standard forestablishing probable cause for electronic surveillance warrants asrequired for more traditional warrants.122 Thus, it is likely thatall twenty-five surveillance requests were supported by sufficientprobable cause.The judicial application of Section 2518 (3) presents a moreelusive question. Although the statutory language expresslymandates that normal investigative procedures generally be triedand failed before electronic surveillance should be authorized, inmany jurisdictions, evidence strongly suggests this procedure isnot adhered to. Some courts authorize electronic wiretappingbefore all other investigative techniques have been exhausted. Inthese jurisdictions, in order to obtain a warrant authorizinginterception of such communications, the government need onlyshow that other techniques would be “impracticable under thecircumstances.”123 In other jurisdictions, law enforcement only119. 18 U.S.C. § 2518 (1994 & Supp. V 1999).120. Id.; see also Dempsey, supra note 46, at 85.121. See Hearings 2, supra note 22 (statement of Donald Kerr, Asst. Director,Federal Bureau of Investigation).122. See generally United States v. Clements, 588 F2d 1030 (1979) (findingprobable cause existed based on information provided by three informants andindependently corroborated by police); United States v. Wagner, 989 F.2d 69 (1993)(finding probable cause existed based on information provided by informant that hadmade eight previous drug purchases under police supervision).123. United States v. Cooper, 85 F. Supp. 2d 1 (2000).

has to state a “likelihood of failure” of other investigativetechniques in their affidavit for a search warrant.124This trend is also reflected statistically. Between 1968 and1995, federal and state courts approved 20,107 surveillanceapplications and denied only 27, with none denied since 1988.Further, from the 349 federal taps authorized in 1991, 1022persons were arrested and 292, or about .08% of the total numberunder surveillance, were convicted.125Thus, based on this persuasive data, it is possible that so longas the FBI can establish probable cause for a particularinvestigation, it can obtain an order to implement Carnivorewithout presenting evidence of an attempt to utilize moreconventional investigative techniques, and still be in fullcompliance with the statutory requirements.Major Crimes OnlyIn drafting the original Title III, Congress recognized thatwiretapping created a significant opportunity for abuse, andtherefore should be limited in its use. Thus, the Legislatureexpressly stated that “interception of . . . communicationsshould . . . be limited to certain major types of offenses and specificcategories of crime with assurances that the interception isjustified and that the information obtained thereby will not bemisused.”126However, the list of offenses for which wiretapping andelectronic surveillance are permitted under Title III and theEPCA has increased from 26 in 1968 to 95 in 1996.127 The list hasexpanded from espionage, treason, violent crimes, and offensestypically associated with organized crime, to include such cases asthose involving false statements on passport applications and loanapplications.128 Further, by and large, electronic surveillance isprimarily used in drug cases. In 1996, 71% of wiretaps nationwidewere issued for drug cases.129124. United States v. Caruso, 415 F. Supp. 847 (1976).125. See Slobogin, supra note 112, at 190.126. Omnibus Control and Safe Streets Act of 1968, Pub. L. No. 90-351, § 801, 82Stat. 211 (1968).127. 18 U.S.C § 2516 (1994); see also Dempsey, supra note 46, at 76-77.128. See Dempsey, supra note 46, at 76.129. Id.

The FBI has stated that its use of Carnivore has compliedwith the “major crimes” requirement of Title III, primarilyfocusing on cases such as those involving the illegal solicitation ofsex with minors and security issues concerning illegal bombmaking activities.130 The Bureau has also acknowledged possibleuse of the device in narcotics investigations.131 In light of theexpanded list of legislatively approved uses for electronicsurveillance, the FBI’s assertion of compliance with the statute iscorrect, and Carnivore is being utilized well within the statutoryguidelines.Minimization of Non-Pertinent InformationThe EPCA requires that electronic surveillance “be conductedin such a way as to minimize the interception of communicationsnot otherwise subject to interception.”132 However, similar to theother two statutorily enumerated protections, the minimizationrequirement has not been strictly enforced by the courts.133 Itappears that any explanation for law enforcement’s failure tominimize will suffice to justify their actions. For instance, in Scottv. United States,134 the D.C. Circuit held that the completerecording of all conversations on a phone line used by a suspectwas acceptable despite the fact that 60% of calls were notpertinent to the investigation. The court accepted the failure tominimize on law enforcement’s explanation that the suspect oftenused coded language in brief conversations. Thus it was necessaryfor law enforcement to listen to all calls to determine theirrelevancy.135Similarly, in United States v. Ozar,136 the Eighth Circuitupheld the FBI’s method of listening to two out of every threeminutes of every phone conversation. In Ozar, the governmentintercepted a total of 8,126 minutes of the suspect’s telephone130. See Qaisar Alam, E-Mail Surveillance: Carnivore Cornered, Computers Today,Oct. 31, 2000, at 48.131. See Danny A. Defenbaugh, FBI’s Carnivore Software, Dallas Morning News,Aug. 27, 2000, at 4J.132. 18 U.S.C. § 2518(5) (1994 & Supp. V 1999).133. See Dempsey, supra note 46, at 77.134. 516 F.2d 751 (D.C. Cir. 1975); see Dempsey, supra note 46, at 77.135. Scott, 516 F.2d at 755.136. 50 F.3d 1440 (8th Cir. 1995); see Dempsey, supra note 46, at 77.

conversations, of which 223 minutes, or 2.75% were deemedpertinent to the investigation.137 The FBI explained that it wasnecessary to listen to a large number of conversations todetermine their relevancy, not because they were short and coded,as in Scott, but because they were lengthy and containedcomplicated subject matter.138The Carnivore system cannot be implemented withoutintercepting electronic transmissions unrelated to the targetsuspect.139 Also, the system operates by intercepting and copyingall communications sent to or received by an individual undersurveillance.140 Thus, a situation requiring effective minimizationof information not pertinent to the investigation is very muchpresented. However, based on the collective reasoning applied bythe judiciary, it appears improbable that a failure by the FBI toadequately minimize the interception of such data whileimplementing Carnivore would have any significant statutoryramifications.Post-Surveillance Judicial ReviewEPCA provides that that an individual who is subject togovernment conducted electronic surveillance is entitled to anafter-the-fact judicial review of the authorization and conduct ofthe surveillance.141 However, the EPCA currently contains noexclusionary provision for illegally obtained electronictransmissions in transit.142 Arguably, this would be the type ofelectronic information Carnivore expressly targets. A proposedamendment to the EPCA suggests the implementation of such a137. Ozar, 50 F.3d at 1448.138. Id.; see also Dempsey, supra note 46, at 77.139. See Hearings 1, supra note 15 (statement of Sen. Patrick Leahy).140. See Hearings 2, supra note 22 (statement of Donald Kerr, Asst. Director,Federal Bureau of Investigation).141. See Digital Telephony and Law Enforcement Access to AdvancedTelecommunications Technologies and Services: Joint Hearings on H.R. 4922 and S.2375 Before the Subcomm. on Tech. and the Law of the Senate Comm. on the Judiciaryand the Subcomm. on Civil and Constitutional Rights of the House Comm. on theJudiciary, 103rd Cong. 6 (1994) (testimony of Louis J. Freeh).142. See Hearings on H.R. 5018, Electronic Communications Privacy Act of 2000;H.R. 4987, Digital Privacy Act of 2000; and H.R. 4908, Notice of Electronic MonitoringAct, 106th Cong. (2000) (statement of Gregory T. Nojeim, Legislative Counsel,American Civil Liberties Union).

provision,143 but judging from the application of exclusionaryprovisions applicable to wire and oral communications, it is notencouraging that such a provision would be effective. Statisticsindicate that a defendant’s after-the-fact challenges to theauthorization or conduct of oral and wire surveillance are rarelysustained.144 Between 1985 and 1994, judges nationwide granted138 suppression motions while denying 3,060, for a 4.3%suppression rate.145 In light of this data, it seems highlyimprobable that even after the enactment of an applicableexclusionary provision, any party attempting to challenge theconduct of an electronic surveillance utilizing Carnivore would besuccessful. From this analysis, it is presumable that the use ofCarnivore would not be deemed to violate the privacy protectionsembodied in Title III and the EPCA.CONCLUSIONThe information and analysis presented in this commentsupports the conclusion that the FBI’s Carnivore surveillancedevice is likely to be in compliance with both constitutional andfederal statutory requirements. However, in light of citizenconcern and uncertainty concerning the utilization of Carnivore, Ibriefly propose two plausible solutions that would likely quell, orat least diminish, these fears and concerns while still achievingthe government objective. These proposals are as follows:First, the FBI could replace Carnivore with a less invasivesurveillance system. Since the existence of Carnivore was madepublic last year, several software companies have developedalternatives. One such alternative, “Altivore,”146 works much likeCarnivore. However, unlike Carnivore, Altivore allows for thecollection of just one stream of data as the information flowsthrough the device, thus reducing some fear that unauthorizeddata is being intercepted.147143. See Hearings on H.R. 5018, Electronic Communications Privacy Act of 2000;H.R. 4987, Digital Privacy Act of 2000; and H.R. 4908, Notice of Electronic MonitoringAct, 106th Cong. (2000) (statement of Robert Corne-Revere).144. See Robert Plotkin, Breaking the Code: Excluding Illegal Wiretap Evidence, 10BNA Crim. Prac. Manual 432 (1996).145. See Dempsey, supra note 46, at 77.146. See Ann Harrison, Security Software Vendor Develops Carnivore E-MailMonitoring Alternative, Infoworld Daily News, Sept. 21, 2000.147. See id.

Second, ISPs should be given some collaborative role in theconduct of the surveillance. Carnivore, in departure fromordinary telephone taps, inserts the FBI into the ISP’s network.148In fact, the ISP has no role in the surveillance once the ISP assiststhe FBI in connecting to its network.149 Like CALEA, the EPCAshould be amended to require ISP assistance in the collection ofdata from its network. In addition, the FBI should make thetechnology of Carnivore, including the source code and the right tomodify it, available to any ISP that needs to comply with asurveillance order. The involvement of the ISP would serve toconfirm exactly what the operating capabilities of the systemreally are.Obviously, these recommendations are not all of the possiblesolutions that could effectively balance the interest of Internetusers with that of the Government. Yet, whatever resolution isselected, American society must act quickly if it wishes to preservefundamental Fourth Amendment protections as it moves furtherand further into the technology age. Without a doubt, in light ofthe uncertain and inadequate legal and judicial guidancecurrently applicable to the Internet forum, we have a very longway to go.Patricia K. Holmes148. See Dempsey, supra note 19, at AO3.149. See id.; see also Hearings 1, supra note 15 (statement of Sen. Patrick Leahy).