Conclusions and Recommendations
Three technological trendsthe ubiquity of information in digital form, the widespread use of computer networks, and the rapid proliferation of the World Wide Webhave profound implications for the way intellectual property (IP) is created, distributed, and accessed by virtually every sector of society. The stakes are high in terms of both ideology and economics. Not surprisingly, much discussion of these issues has occurred in the Congress, among stakeholder groups, and in the press. But the effects of the information infrastructure extend beyond these institutions; as never before there are also important and direct effects on individuals in their daily life.
The information infrastructure offers both promise and peril: promise in the form of extraordinary ease of access to a vast array of information, and peril from opportunities both for information to be reproduced inappropriately and for information access to be controlled in new and problematic ways. Providing an appropriate level of access to digital IP is central to realizing the promise of the information infrastructure. Ensuring that this appropriate level of access becomes a reality raises a number of difficult issues that in the aggregate constitute the digital dilemma. This report articulates these difficult issues, provides a framework for thinking about them, and offers ways of moving toward resolving the dilemma.
One salient theme in the committee's conclusions and recommendations is an acknowledgment of the multiplicity of stakeholders and forces that must be considered. Intellectual property has a pervasive impact in
society, resulting in a corresponding diversity of interests, motivations, and values. Some stakeholders see the issues in economic terms; some in philosophical terms; others in technological terms; and still others in legal, ethical, or social policy terms. There are also a variety of important forces at workregulations, markets, social norms, and technologyall of which must be considered and all of which may also be used in dealing with the issues. Knowing about the full range of forces may open up additional routes for dealing with issues; not every problem need be legislated (or priced) into submission. Individuals exploring these issues are well advised to be cognizant of all the forces at work, to avoid being blind-sided by any of them; to avail themselves of the opportunity to use any of the forces when appropriate; to be aware of the process by which each of them comes about; and to consider the degree of public scrutiny of the values embedded in each.
The committee believes that the issue of intellectual property in the information infrastructure cannot be viewed as solely a legal issue (as it was, for example, in the white paper Intellectual Property and the National Information Infrastructure, IITF, 1995)1 or through any other single lens. Such an approach will necessarily yield incomplete, and often incorrect, answers. One of the committee's key contributions is to urge an appropriately broad framework for use by policy makers, one that acknowledges the full spectrum of stakeholders and forces.
The first two sections of this chapter focus on the implications for society and individuals that arise from the everyday use of the information infrastructure, with an emphasis on intellectual property that has been published in the traditional sense.2 The next two sections address research and data collection that are needed and near-term actions that can be initiated to help in getting beyond the digital dilemma. The last section offers guidance on and principles for the formulation of law and public policy.
A significant portion of the committee's deliberations can be characterized as spirited and energetic discussions expressing a range of perspectives on controversial issues. For some of those issues, a summary of alternative perspectives is provided, with the intent of exposing the core issues to aid future discussion. That this committee, a diverse and balanced group of experts, had difficulty in achieving consensus in many areas, despite extensive briefings, background reading, and deliberations,
1When the IITF's white paper was written, the Web was only beginning to be widely used by the general public; hence some aspects of the digital dilemma touched on here (e.g., business models) had yet to develop.
2The committee was unable to address some important subjects (e.g., the cable television industry) thoroughly because of the limited time and resources available.
should serve as a caution to policy makers to contemplate changes to law or policy with the utmost care.
The Digital Dilemma: Implications for Public Access
Public access to published works is an important goal of copyright law. The traditional model of publicationthe distribution of physical copies of a workhas been effective as the fundamental enabler of public access.3 Enough copies of a work are usually purchased (e.g., through libraries and other institutions and by private individuals) that it becomes part of the social, cultural, and intellectual record and is thus accessible to sufficiently motivated members of the public. There is also a long-standing (if not always explicitly articulated) understanding that this social and cultural record will continue to accumulate, be preserved, and be available for consultation. At least since the modern era of public libraries, broad access to a college education, and mass media, such information has become increasingly available. Yet there are aspects of the information infrastructure that, although vastly increasing access in some ways, also have the potential to diminish that access, which is a valuable component of our social structure.
The Value of Public Access
Public access, and the social benefits that arise from it, may be an undervalued aspect of our current social processes and mechanisms. As one example, while the first-sale rule enables access that may result in loss of revenue for publishers (because some people or organizations who are able to borrow a book would have purchased it instead),4 the larger social benefitsan informed citizenry and the democratization of information and knowledgecan be substantial.5 Those benefits also have a significant and longer-term impact in encouraging the creation of new knowl-
3The traditional model of publication has been more applicable to some forms of information (e.g., books and magazines) than others (e.g., first-run movies and television broadcasts).
4The first-sale rule stipulates that the initial sale of a copy of a work exhausts the copyright owner's right to control further distribution of that copy. An individual, library, or other entity is free to give away, lend, rent, or sell its copies of books and many other materials (17 U.S.C. sec. 109).
5Some materials (e.g., academic journals) are rarely bought by individuals and hence would not represent any substantial lost revenue. There may also be some countervailing effect, because some people who get access to a book through borrowing are motivated to buy it; lending is in effect a form of advertising. The point here is that even if there is some degree of loss, the benefits must also be considered.
edge and new works. Being well informed and educated has value that increases with the population of others similarly informed and educated, and ultimately contributes to a larger potential market for authors and publishers. Hence, the public access to material that is made possible in the hard-copy world by the first-sale rule can be worth more to society than the modest revenue lost to publishers. Beyond the economic issues, an informed citizenry and informed discourse are vital to the health of a free and democratic society.
Public access may suffer, however, as the evolution of the information infrastructure compels a reexamination of the first-sale rule and other mechanisms for achieving access. As one example of the difficulties digital information brings, a single online copy of a work available from a digital library could diminish the market for the work much more than the distribution of hard copies to traditional libraries. One reasonable response of publishers might be to avoid making some works available to libraries in digital form, resulting in a net decrease in the accessibility of information. Other challenges to ensuring access arise from the changing nature of publication, the growing use of licenses rather than sale of works, and the use of technical protection mechanisms. As a consequence, historically simple provisions such as the first-sale rule become much more complex in the digital environment, involving difficult questions with respect to technology and business practices.
Conclusion: The tradition of providing for a limited degree of access to published materials that was established in the world of physical artifacts must be continued in the digital context. But the mechanisms for achieving this access and the definition of ''limited degree" will need to evolve in response to the attributes of digital intellectual property and the information infrastructure.
Consequences of the Changing Nature of Publication and the Use of Licensing and Technical Protection Services
In liberating content from its medium of presentation, digital information challenges many long-held assumptions about copyrighted works, most notably those regarding the nature and character of publication. In the physical world publication is public, irrevocable, and provides a fixed copy of the work; in the digital world none of these may be true.
Publication has traditionally been public in the sense noted above (i.e., that works are widely distributed and become part of the cultural record). Publication is irrevocable in the sense that works may go out of print, but once published can never subsequently be effectively with-
drawn from circulation and become "unpublished." Publication also implies a stability of the work. Copies distributed provide a stable snapshot of the work at a particular moment; subsequent editions only add to this record. In the digital world, however, documents published by being posted on the public Internet can be removed from scrutiny at the pleasure of the rights holder. Access can be controlled to allow many levels of dissemination between publication and private distribution, and older versions of a document can be (and are routinely) replaced by newer ones, obliterating any historical record.
The widespread use of licensing and technical protection services (TPSs) also has important implications. Licensing is a familiar mechanism for providing access to some types of digital information (e.g., software) but is relatively new for other types (e.g., research journals). Even where the practice is familiar, it has often stirred controversy, as in the still-developing notion of shrink-wrap licenses. Where licensing is unfamiliar, publishers and their customers are still learning how to establish reasonable licensing relationships. By offering a distribution model different from that represented by copyright and sale, licensing has the potential to open new markets. Some material that has been made available through licensing would not have been published at all in the traditional manner; the restricted distribution of information is thus an important option for the publisher and public.
But the use of licensing also raises significant concerns about the consequences for public access and the maintenance of a healthy corpus of materials in the public domain, particularly where license restrictions differ from legal rules that would otherwise apply. The libraries' role as a permanent repository of material that constitutes a cultural heritage is threatened by a change in the model of distribution from sale to licensing. Libraries could instead become transient, temporary points of access to collections of information that may be available today and gone tomorrow, when licenses expire. Additional concerns arise from the fact that material distributed by license may not become a part of the long-term public record.
Some technical protection services have been developed (and others are being developed) to confront the key problem that digital information seemingly cannot be distributed without the risk of large-scale copying and redistribution.6 TPSs offer rights holders some assurance that distributing a single copy of a digital work need not result in subsequent unlimited and uncontrollable dissemination. By enabling network distri-
6Technical protection services are discussed further below under "Moving Beyond the Digital Dilemma: Additional Mechanisms for Making Progress" and in greater detail in Chapter 5.
bution of information products that otherwise would never have been distributed digitally, TPSs could open new markets and substantially increase dissemination of and access to works. Conversely, without such an ability, some rights holders may decide to avoid digital distribution entirely for some works (e.g., investment newsletters), thereby reducing the offerings available to the general public.
But technical protection services may also permit limitations on the distribution of content such that most consumers can only view itdistribution without the ability to save and/or print is now contemplated for several mass market content businesses. Consider the consequences of this model of publication: Information might be distributed but never easily shared, substantially defeating the original intent of publication as an act that leads, eventually, to a contribution to the shared, permanent social and cultural heritage. Time- and audience-limited distribution could increase.7
Although limited distribution is a possibility, it may prove to be thoroughly unpopular with consumers if it significantly constrains access to, enjoyment, or use of a product. The marketplace might thus facilitate public access. Nevertheless, policy makers should monitor the situation and be prepared to address the issue in the event that limited distribution models begin to have a significant impact on public access to information.
Conclusion: The confluence of three developmentsthe changing nature of publication in the digital world, the increasing use of licensing rather than sale, and the use of technical protection servicescreates unprecedented opportunities for individuals to access information in improved and novel ways, but also could have a negative impact on public access to information. Developments over time should be monitored closely.
Some members of the committee voiced the concern that highly constrained models of distribution undermine the fundamental pact between society and authors that is embodied in copyright, a pact that encourages the creation and dissemination of information for society's ultimate benefit. These individuals are concerned that a limited-distribution model of publication may undermine a constitutional intent, namely that rights be granted to authors for a limited time in exchange for assurance that materials will pass eventually into the public domain and the public record.
7Time- and audience-limited access has been commonplace for some kinds of IP for many years (e.g., movies exhibited in a theater), but this is a new phenomenon for traditionally published IP.
Recommendation: Representatives from government, rights holders, publishers, libraries and other cultural heritage institutions, the public, and technology providers should convene to begin a discussion of models for public access to information that are mutually workable in the context of the widespread use of licensing and technical protection services.
Publication and Private Distribution
In a digital world offering options for distribution other than printing and selling copies, it is not always easy to tell when information has been published and when it has not. The distinction between publication and private distribution is blurred by options such as distribution on electronic mailing lists, posting on password-protected Web sites, or posting on preprint servers available to members of professional societies. Further blurring results from the multiple, finely controlled layers of conditional access that computer systems can provide, offering many degrees of access between public and private. The issue is further complicated by the impermanent nature of digital information, which facilitates the distribution of works in varying states of completion (e.g., posting numerous versions of an article as it evolves). The question of what constitutes publication has significant consequences with respect to public access to the informationfacts and ideas in published works are freely available to the publicbut deciding whether a work in digital format has been published may be difficult. Although the distinction between public and private may never have been crystal clear in the copyright regime, it has become far murkier in the digital environment.
Conclusion: The information infrastructure blurs the distinction between publication and private distribution.
Recommendation: The concept of publication should be reevaluated and clarified (or reconceptualized) by the various stakeholder groups in response to the fundamental changes caused by the information infrastructure. The public policy implications of a new concept of publication should also be determined.
Mass Market Licenses
Non-negotiated licenses for mass market items also raise important public access questions. The issue is whether the terms of mass market licenses offered on a take-it-or-leave-it basis would override fair use or
other limiting policies of copyright law. The question is controversial and as yet unresolved in the law. The public policies associated with intellectual property law may sometimes be seen as sufficiently important that mass market license terms should not be permitted to override them. For example, public policy favoring competition and innovation may call into question the enforceability of a term in a mass market license for computer software that forbids reverse engineering the software. Similarly, concerns related to free speech may arise if mass market licenses seek to limit criticism of a digital information product or disclosure of its flaws. Part of the intent of fair use is to encourage critical analysis; however, if works are licensed, there is currently no automatic fair use provision and hence no established foundation for criticism.8
Some committee members favor subjecting mass market licenses to fair use limitations, viewing fair use and other limiting doctrines of copyright as having an affirmative characteri.e., as providing a right for users under copyright law, rather than solely a defense to infringement. According to this view, rescinding that right in a license should not be possible (even though other rights may, with few exceptions, be waived by agreement). Those who do not favor subjecting mass market licenses to fair use conditions generally perceive copyright as providing default rules that should be overridable by a contract in free market transactions.
Conclusion: The committee as a whole points out an important underlying legal and philosophical issuethe question of whether fair use is an affirmative right or a defenseand emphasizes the consequences for access that follow from taking one position or the other on this issue.9
Archiving and Preservation of Digital Information
Archiving our cultural heritage and ensuring a record of intellectual discourse are critical tasks for society. The importance of archiving is discussed in Chapter 3, along with many of the associated problems.
8Consider the hypothetical case of an electronic commerce software package and an authorized user who discovers a security problem with the software. The vendor may wish to issue licenses that prohibit the authorized user from disclosing such information to third parties.
9Similarly, the committee is unable to take a definitive position with respect to the Uniform Computer Information Transactions Act (UCITA). See Chapter 3 for a discussion of the issues concerning UCITA.
Despite several years of intense effort, little practical progress is being made in establishing digital archives. The lack of progress is attributable to several factors:
• Lack of funding for large-scale digital archiving overall and lack of any agreement about how responsibility for providing funding will be divided among government and cultural heritage research institutions. There is also an absence of mechanisms to effectively pool the contributions of the many organizations with some interest in and responsibility for funding archiving.
• Reluctance among major research libraries and archives to make digital archiving a part of their missions or, if there is acceptance of responsibility, the inability to proceed without certainty of funding. Justification of funding is complicated by the difficulty of offering any real access to materials prior to the expiration of copyright, with the result that digital archives may not produce tangible benefits for a century or more, making this investment in the preservation of culture and scholarship a hard sell.
• Insufficient expertise within the most likely archiving institutions; the technical and intellectual problems involved are difficult and experienced individuals correspondingly difficult to find.
• Worry about potential liability for contributing to copyright infringementsfor example, fear that any unauthorized use of archived material by a member of the public could result in the archive being held liable for contributing to infringement. Archives are also concerned about liability for copyright infringement, both in the actual processes of capture and management of digital content and in any actions taken to make archived digital materials available to the public prior to the expiration of the term of copyright.
• The daunting scale of the task and the need to develop processes for selecting what will be preserved. Hard intellectual and technological problems exist, some of which require the development of social and scholarly consensus.
• The uncertain relationship between archiving and licensing. The rights to archiving can be negotiated, and indeed many research libraries are starting to do so. These negotiations seem to have been reasonably successful thus far when carried out with scholarly publishers that share an interest with libraries, authors, and readers in ensuring that electronic publications will be archived. The likelihood of success is less clear with mass market publishers and content providers outside the print tradition (e.g., in the music industry). Licensing is simply a contract between a publisher and a client, so the publisher is under no obligation to include provisions for archiving. This situation is unlike that in the print world,
where publishing and selling a book (for instance) automatically make archival preservation possible both logistically and legally.
There are significant economic and legal issues to be resolved if archives and libraries are to act as digital archives during the term of copyright. Acquiring works after copyrights expire is clearly ineffective, as it is unlikely that most works will be available for acquisition: Few digital (or traditional) works remain economically viable for 100 years, and thus available in the marketplace.
The Congress, the Administration, and the combined managements of the top research libraries and archives should lead these efforts. Preservation of works on paper and in other physical media has evolved as a responsibility shared by many autonomous institutions. The committee advocates a similar approach to digital archives and encourages practical steps toward distributed digital archives for which existing research libraries and archives share responsibility. The committee believes that starting now is urgent, so that digital archives of significant extent will be established within a decade. The first step is to initiate a process that engages all relevant stakeholders, develop a plan for moving forward, and begin assembling the political constituency that will ultimately be needed to implement the necessary actions.
Conclusion: Significant economic, technical, and legal issues need to be resolved if libraries and archiving institutions are to be as successful with digital information as they have been with hard-copy information.
Recommendation: A task force on electronic deposit should be chartered to determine the desirability, feasibility, shape, and funding requirements of a system for the deposit of digital files in multiple depositories. The task force membership should broadly represent the relevant stakeholders and should be organized by an unbiased entity with a national reputation, such as the Library of Congress or some other governmental organization that has a pertinent charter and relevant expertise. The task force should be assigned for a limited term (2 years maximum) and should be charged with the following responsibilities:
• Determining the desirability, feasibility and general design of a system for the deposit of digital files in multiple depositories;
• Considering incentives for rights holders that encourage
their participation in such a system and encourage them to move material into the public domain earlier than the term of copyright protection if that material is not being exploited;
• Proposing both the legal and procedural framework for the deposit and subsequent use of digital files;
• Addressing the intellectual property and liability concerns of libraries and rights holders;
• Reporting on and recommending to Congress the long-term funding requirements for making the system work. A recommendation on funding is particularly important, because without funding nothing will happen when the task force's study is completed. This committee does not have sufficient information to indicate what the congressional or total funding level should be, but wants to make clear its belief that the total funding needed is substantial; and
• Evaluating other nations' strategies for the deposit and preservation of information and considering how a U.S. system could build on and relate to these other national efforts, recognizing that the creation and dissemination of digital information are global activities, and that preservation of content is thus a global problem.10
Preservation within the context of the information infrastructure introduces new challenges. Digital information is often stored on media with relatively short life spans: because the medium itself degrades (e.g., magnetic tape) or the relentless advance of formats, hardware, and soft-
10To illustrate possible outcomes of the task force, one scenario might call for voluntary (or mandatory) deposit of digital works that are protected by copyright in the United States and that are either offered for sale under license or, if distributed free of charge, are protected by a TPS. Such deposited copies would not be made available to the public by the depository as long as they are still offered to the public by the rights holder, except for viewing in the library itself (as is the case with hard-copy works). All deposited copies would be "in the clear" (i.e., with no encryption or other access-limiting mechanism). The depositors would have no technical responsibility for migrating the copies over time. Libraries and other archives would not be held liable for unauthorized access to these files but would be required to take reasonable steps to prevent and stop such violations.
The intent here is to extend into the digital world the traditional balancing act of IPproviding enough control over a work to offer an incentive for creation, yet ensuring that in the long term all work becomes a part of the public intellectual record to the benefit of society as a whole. Providing for deposit of materials "in the clear" may aid in dealing with problems of access that arise from technical protection mechanisms, as well as issues raised by archiving.
ware leaves behind orphan technologies (try finding a way to read old word processor files on an 8-inch floppy disk). As a result, preserving digital information inevitably involves copying it to a new medium and/ or format. If digital information is to be preserved, such copying will need to be permitted under the law. In some instances, literal copying may not be practical or desirable, and the information may need to be adapted for a new format. This migrating process goes beyond simple copying and might be construed as the creation of a derivative work, which will also need to be allowed under the law to enable preservation.
Technical protection services may also pose difficulties for migrating or accessing archived digital information. As far as the committee was able to ascertain, no TPS incorporates any kind of "self-destruct" feature that is triggered when a work legally enters the public domain; rather, a TPS will continue to try to control access forever. In addition, a TPS may try to block attempts to copy or reformat a work, even when such a step is legal as part of the archival management of content. A self-destruct feature may be technically feasible and could be incorporated in future products and services. Access problems could also arise from TPSs that require online authorization or other interaction with network-based servers before they will permit access: What happens if the publisher ceases support of the authorization system, or ceases to exist at all11
Maintaining digital documents in the face of changing hardware and software over even a decade can be a challenging prospect; yet, under certain circumstances, works can remain protected by copyright for more than 100 years after their date of creation. What incentive is there for the publisher to maintain access to a century-old work (which may mean reformatting it, updating the TPS that surrounds it, and so on), especially if it is providing no appreciable income?
The digital preservation problem is becoming a consumer issue as well. Consumers own content in media and formats that are becoming obsoletefor example, vinyl long-playing (LP) record collections today and perhaps audio CDs in the future. Should consumers legally be able to migrate content that they have purchased to new technologies for their own use?
Recommendation: Congress should enact legislation to permit copying of digital information for archival purposes, whether the copy is in the same format or migrated to a new format.
11A recent example is the discontinuance of Divx digital video service.
Access to Federal Government Information12
Advances in the capabilities and use of the information infrastructure, most notably the widespread use of the Web, have provided the means for greatly expanded access to federal government information. Agencies are increasingly using the Web to make information available, with some notable successes, including the Government Printing Office's GPO Access system, the Library of Congress's THOMAS system, and the Security and Exchange Commission's EDGAR system.
However, in some parts of the government, the evolution of the information infrastructure has instead been associated with a trend toward the commercialization of government information, increasingly limiting the amounts of information that can be accessed inexpensively by the public. Broad access to and use of publicly funded information are inhibited when distribution agreements curtail the availability of information. In some cases, federal agencies themselves must pay high prices for data or products that were created by another federal agency.
In addition, some agencies that perform research and development may contract with private companies for commercial ventures involving the results of their research. Such relationships can produce incentives to be less open in sharing research results. Other agencies, such as those involved with managing the records of government, may similarly put a low priority on public access and have, in some cases, allowed private parties to copy their internal records for the purpose of packaging and reselling the resulting data products.
Conclusion: When commercial enterprises add value to basic data, the resulting products deserve copyright protection insofar as these products otherwise satisfy the legal requirements for copyright.
Recommendation: As a general principle, the basic data created or collected by the federal government should be available at a modest cost, usually not to exceed the direct costs associated with distribution of the data. When agencies contract with a commercial enterprise to make federally supported primary data available, and provide no other mechanism for access to the data, such agreements should provide for public access at a
12The committee recognizes that state and local governments, as well as the governments of other nations, produce and distribute valuable data and information and face many of the same issues as the federal government regarding the dissemination of information. Because of the limited time and resources available, the committee focused its deliberations on data and information produced and distributed by the federal government.
cost that does not exceed the direct costs associated with distribution.13
The Digital Dilemma: Implications for Individual Behavior
Perceptions and Behavior of Individuals
Little is known about how frequently individuals duplicate copyrighted materials and whether they pause to question whether this activity may be illegal. It is likely that a large number of people assume that they have the right to duplicate such material and that their view of appropriate conduct is not shaped by any substantive knowledge of intellectual property law. For the past decade or more, most individuals who photocopied books and journals in a library have encountered signs warning them about potential copyright infringement, but little is known about whether these signs have been seen or understood, or have resulted in a change in user behavior. Individuals attempting to copy videotapes are confronted with similar but more threatening on-screen warnings at the beginning of a tape, but little is known about how these warnings have affected behavior. Previous studies, some more than a decade old, have typically examined behaviors and attitudes of only narrow groups of people prior to the widespread use of the Internet.14
Anecdotal evidence suggests that most people are not generally informed about copyright in the context of the information infrastructure; instead, myths and misunderstandings abound regarding what is legal and what is not. The committee believes that such misunderstandings extend to contracting arrangements as well. Few people read and understand shrink-wrap or point-and-click licenses, and whether people think they need to take them seriously is unclear.
Conclusion: A better understanding is needed of the public's perception and behavior concerning digital intellectual property. When popular attitudes and practices are out of synch with laws, the enforcement of laws becomes more difficult, which may instill in people a lack of confidence in and respect
13The committee did not address the status of the data and research created by federally supported researchers based at academic or other institutions outside the federal government.
14The most significant study thus far, done by the former congressional Office of Technology Assessment (1989), focused primarily on the home recording of audiotape and briefly examined home video recording as well. Research that is more relevant and current is needed.
for the legal system. There are also political dangers associated with criminalizing generally accepted behavior, given the possibilities for discriminatory and selective enforcement.
Recommendation: Research and data collection should be pursued to develop a better understanding of what types of digital copying people think are permissible, what they regard as infringements, and what falls into murky ill-defined areas. Such research should address how these views differ from one community to another, how they differ according to type of material (e.g., software, recorded music, online documents), how user behavior follows user beliefs, and to what extent further knowledge about copyright law is likely to change user behavior.
Fair Use and Private Use Copying
As discussed in Chapter 4, fair use is an established doctrine of U.S. copyright law that has become especially controversial in recent years. One controversy concerns the extent to which private use copying of copyrighted materials can be justified as fair use. Although this issue is not unique to digital intellectual property, the ease with which digital copies can be made and distributed, especially in networked environments, makes private use copying far more extensive in the digital environment and a more significant problem for content owners. A second controversy concerns the viability of fair use and other limitations on copyright in the digital environment.
The end points on the spectrum of perspectives on this issue are that private use copying is almost always fair use as a matter of copyright law and that private use copying is never fair use. Neither position is correct, although the committee concluded that many members of the general public consider the former to be true. Private use copying is sometimes fair use, and it is sometimes illegal; as a separate matter, it is sometimes ethical and sometimes not. Although it may be difficult to accomplish in practice, whether a private use is a fair use should in principle be determined by considering the fair use provision, section 107, of the copyright law (see Chapter 4).
The committee identified several considerations that complicate articulating a specific position on the copying of information for private use. One difficulty is the lack of clarity about what ''private" means in the context of copying (e.g., copying in one's home may be "private," but what about copying in a library or a school?). Another is the multifaceted nature of fair use determinations, which makes it difficult to articulate
simple, straightforward rules that could guide the conduct of the average citizen. In addition, the private nature of such copying poses serious enforceability problems, and, for some, privacy and other social values may make stricter enforcement socially unacceptable or undesirable. Nevertheless, although enforcement of regulations concerning private use is clearly not easy, the difficulties of enforcing the law do not transform private uses into fair uses when other considerations suggest that they are not.
Conclusion: A widespread (and incorrect) belief prevails in society that private use copying is always or almost always lawful. This viewpoint is difficult to support on either legal or ethical grounds. It is important to find ways to convince the public to consider thoughtfully the legality, ethics, and economic implications of their acts of private copying.
What can or should be done concerning this widespread and incorrect belief? The committee identified three distinct views and corresponding actions:
• One view is that copyright law should not concern itself at all with private use copyingthat, as a pragmatic position, such copying should be considered lawful. Because private use copying cannot effectively be controlled at reasonable cost and without impeding other important social values, branding it unlawful might encourage a more general flouting of intellectual property law.
• A second view is that much private use copying is fair use and that illegal copying could be discouraged by appropriate copyright education.
• A third view sees most private use copying as illegal under copyright law and suggests that more significant efforts should be undertaken to enforce rules against illegal private use copying. This group would also advocate an extensive copyright education campaign specifically directed against private use copying. Technical protection services may also make private use copying more controllable than it currently is.
As in the print world, in the digital world basic differences in values make it difficult to determine the proper boundaries of fair use. Different stakeholders weigh the values and interests reflected in copyright law and policy differently. Those who find private use copying illegal generally do so on the grounds that the copyright law's exclusive right to reproduce is indeed truly exclusive and does not restrict only copies made for public distribution. They believe that privately made copies are no less an infringement for being private. Those who would hold private use
copying to be fair use generally refer to the traditional standards for fair use, notably the claimed lack of impact on the right's holders market for a protected work.
Beyond the issue of private use copying, the committee considered the broader viability of fair use in the digital environment. One rationale for fair use (historically and in the digital environment) has been that transaction costs can make licensing of some copyrighted works uneconomical, leading to failure to use a work even when both the rights holder and potential licensee would agree to such use. The development of the infrastructure for electronic commerce may lower transaction costs for some classes of works. But the committee concluded that the potential for market failure is not the only rationale for fair use or for other exceptions and limitations to copyright law. Other rationales derive from public policy concerns about fundamental human rights, such as freedom of the press, and certain articulated public interests such as preservation of cultural heritage.
Conclusion: Fair use and other exceptions to copyright law derive from the fundamental purpose of copyright law and the concomitant balancing of competing interests among stakeholder groups. Although the evolving information infrastructure changes the processes by which fair use and other exceptions to copyright are achieved, it does not challenge the underlying public policy motivations. Thus, fair use and other exceptions to copyright law should continue to play a role in the digital environment.
Conclusion: Providing additional statutory limitations on copyright and/or additional statutory protections may be necessary over time to adapt copyright appropriately to the digital environment. The fair use doctrine may also prove useful as a flexible mechanism for adapting copyright to the digital environment.
Recommendation: Legal, economic, and public policy research should be undertaken to help determine the extent to which fair use and other exceptions and limitations to copyright should apply in the digital environment. As public policy research, legal developments, and the marketplace shape the scope of fair use and other limitations on copyright, and/or demonstrate a need for additional protections, any additional actions that may be needed to adapt the law, educate the public about it, or enforce the law may become clearer.
With the expansion of the information infrastructure into everyday life and the widespread acknowledgment of the information revolution's power to transform society, the role of information looms ever larger. Yet the committee believes that the public is not well informed about intellectual property law in general and also that it labors under misconceptions concerning copyright in particular. Because ignorance regarding copyright law, the fundamental philosophy it embodies, and its intent may be a significant factor contributing to misuse of protected material, the committee believes that a copyright education program may prove quite useful.
Promoting respect for copyright in the United States would lay an important foundation by educating society about some of the ground rules on which an information-based society is built and help ensure that inadvertent violations of law are not commonplace events. This process would help maintain the health of the information industries (e.g., software, entertainment) and their contribution to the economy, as well as help content creators understand their rights. Respect for and enforcement of copyright law in the United States would also provide a foundation for U.S. efforts aimed at enforcement of international agreements on IP.
The committee believes that, to be effective, a program of copyright education must clearly communicate that the law is, in its intent and spirit, attempting a fundamentally fair and equitable balancing of interests. The program should emphasize the core goal of IP law, namely, the improvement of society through the advancement of knowledge; should describe the difficult balance between control and dissemination; and should make clear that, in the long term, all intellectual property becomes a part of the shared intellectual heritage, available to everyone. Such a program would describe both the rights granted exclusively to creators and the limits on those rights. The program should include an introduction to fair use and other limitations on copyright law, and their role in accomplishing the larger purpose of the law.
Several factors make carrying out successful copyright education difficult, including the complexity of copyright law, disagreements within the population on some significant copyright issues (e.g., the proper scope of private use copying), varying cultural views on the different classes of information,15 and the perceived urgency of other matters on the educational agenda (e.g., can children read?). Appropriate educational materi-
15For example, some people believe that it is more acceptable to copy popular music files in MP3 format for their friends than to copy a word processing software package, even though both acts may be illegal.
als and trained personnel are also lacking, as is empirical research on who infringes and why. All these factors impede development of effective messages for different parts of the population.16
Conclusion: A better understanding of the basic principles of copyright law would lead to greater respect for this law and greater willingness to abide by it, as well as produce a more informed public better able to engage in discussions about intellectual property and public policy.
Recommendation: An educational program should be undertaken that emphasizes the benefits that copyright law provides to all parties. Such a copyright education program needs to be planned and executed with care. Appendix F discusses the rationale for and the desirable characteristics of copyright education.
The committee could not decide how extensive copyright education should be, who should conduct this education, or who should pay for it. However, the committee agreed that copyright education should focus on the basic fairness of the copyright law, should not be oversimplified, and should not be mandated by the federal government.
Moving Beyond the Digital Dilemma: Additional Mechanisms for Making Progress
The difficulties posed by the digital dilemma are formidable. However, progress can be made, and, indeed, several avenues for moving forward have already been discussed in this chapterestablishing a Task Force on Electronic Deposit, recommending changes to the law to facilitate the maintenance of digital archives, encouraging stakeholders to work together to develop mutually agreeable public access models for licensing and technical protection services, and increasing the use of copyright education. Additional means for progress are discussed in this section; the following section discusses the need for research and improved data.
Technical Protection Services
Although technical protection services cannot resolve legal, social, or economic issues underlying intellectual property, they can help to enforce
16This is one reason the committee recommends research to obtain information on public understanding of copyright law, especially for digital intellectual property.
agreed-upon rights, rules, constraints, and responsibilities. Technical protection for intellectual property can play a variety of roles, from helping rights holders collect revenue, to helping safeguard user privacy, to helping ensure information authenticity.
Like any security system, a TPS cannot protect perfectly. Even state-of-the-art systems can be compromised by a sufficiently knowledgeable and determined adversary (who may simply avoid picking the lock by finding ways around it). Hence, with the exception of situations in which security is the overriding concern, TPS design always involves a trade-off between capability and cost, including the cost of the effort of the content distributor (who must use and maintain the system) and the effort of users, who typically experience inconvenience in dealing with the system.17. While this trade-off often results in a distributor's choosing a TPS of only moderate strength, such a solution is frequently entirely adequate and appropriate.
Conclusion: Technical protection services need not be perfect to be useful. Most people are not technically knowledgeable enough to defeat even moderately sophisticated systems and, in any case, are law-abiding citizens rather than determined adversaries. TPSs with what might be called "curb-high deterrence"systems that can be circumvented by a knowledgeable personare sufficient in many instances. They can deter the average user from engaging in illegal behavior and may deter those who may be ignorant about some aspects of the law by causing them to think carefully about the appropriateness of their copying. Simply put, TPSs can help to keep honest people honest.
Conclusion: Technical protection technologies are currently deployed to varying degrees. Some, such as encryption and password protection, are widely deployed. Others, such as Web monitoring, watermarking, time stamping, and rights-management languages, are well developed but not yet widely deployed. Copy prevention techniques are deployed to a limited degree. The copy prevention mechanism used in digital video disks provides a notable example of mature development and consumer market penetration.18
17For a detailed discussion of these costs see Trust in Cyberspace (CSTB, 1999c).
18As noted in Chapter 5, the content scrambling system used in DVDs was cracked in November 1999, illustrating that no system is perfect and that determined adversaries can find ways around state-of-the-art systems.
Concern about computer security for electronic commerce is becoming an increasingly significant influence on the design of computing and communications infrastructure. This could lead to the development and widespread adoption of hardware-based, technologically comprehensive, end-to-end systems that facilitate creation and control of digital IPso-called trusted systems.19 While the installed base of general-purpose PCs represents a large obstacle for trusted systems, efforts to develop standards for secure PC hardware and operating systems for electronic commerce could provide the needed foundation.20 Trusted systems for intellectual property would be a by-product of the efforts to create security for electronic commerce. As of 1999, however, there are no widespread deployments of trusted systems.21
Although TPSs can be based on software alone, protecting valuable content against highly skilled and determined adversaries requires both hardware and software components. Part of the reason is the extent to which software-only circumvention techniques can easily be shared: They are easily distributed worldwide via the Internet, and even unsophisticated users can apply well-designed software (illicit though its purpose may be). Circumvention techniques that require special hardware or hardware-handling expertise are far less easily shared. TPSs with a hardware component are also more effective at making content usable on only one machine, preventing circumvention through redistribution.
But marketplace and infrastructure challenges make developing and deploying specialized hardware difficult except in a few niche markets. Thus, significant attention has been paid to solutions that rely on software only, including some implementations of secure containers and cryptographic envelopes. Several commercial efforts are under way to build and deploy software-based, end-to-end IP delivery systems.
Conclusion: Robust, integrated technical protection services based on the vision of trusted systems using specialized hard-
19The committee uses the phrase "trusted systems" as it is generally understood within the consumer technology market. "Trusted systems" also has a specific and different meaning within the U.S. defense community.
20One such relevant effort is the Trusted Computing Platform Alliance, a collaborative effort founded by Compaq, HP, IBM, Intel and Microsoft. This effort is apparently aimed at just such a goal, trying to provide security at the level of the hardware, BIOS, and operating system. While intended initially as a platform for business-to-business electronic commerce, the technology has clear relevance to protection of intellectual property. For background information, see <http://www.trustedpc.org>.
21The large installed base of personal computers with a fundamentally open design has also created a large market in complementary products (e.g., disks, sound boards, video boards). Because these products would also have to be redesigned to ensure security, they add to the obstacle faced by trusted systems.
ware are not likely to be realized anytime soon for the consumer market. Such systems, if indeed they evolve, will more likely find a home first in the business-to-business electronic commerce sector. TPSs based on software only are being deployed widely, although not without costs: Systems that have been commercialized to date require a substantial infrastructure to manage secure identification of users or authorization of actions.
Conclusion: As cryptography is frequently a crucial enabling technology for technical protection services, continued advances in technical protection services require a productive and leading-edge community of cryptography and security researchers and developers.
Although technical protection services are not yet widely used, some industriesmost notably entertainmenthave made concerted efforts and plan to deploy TPSs to protect their IP (e.g., the Secure Digital Music Initiative discussed in Chapter 2). Industries associated with images (e.g., publishing, news, sports, and entertainment) are adopting watermarking technologies to identify and protect their work. TPSs aimed at text material have begun to appear, but have not yet been widely adopted.
Recommendation: Rights holders might consider using technical protection services to help manage digital intellectual property but should also bear in mind the potential for diminished public access and the costs involved, some of which are imposed on customers and society.22
Not every information product need be distributed by digital networks, given the availability of alternative mechanisms offering most of the advantages and fewer risks. High-value, long-lived products (e.g., classic movies like The Wizard of Oz) might not be available legally on digital networks such as the Internet while protected by copyright, because the consequences of someone capturing the bits are simply too great, and the technical, legal, and social enforcement costs of ensuring that this does not happen are prohibitive. Simply put, the information infrastructure need not be made safe for mass marketing of every form of content.
The pressure to do so is reduced by the possibility of developing
22The enforcement associated with TPSs involves costs that are both private (e.g., enforcement efforts through industry associations) and public (e.g., a pro-rata share of operating the court system).
special-purpose delivery devices (such as DVDs) that combine both software encryption and specialized hardware in a manner that makes the decrypted digital content very difficult to capture. While the specific encryption system used in DVDs was cracked late in 1999, it is still the case that making the content accessible only with specialized hardware can offer substantially more security than is possible with the software-only solutions used when content is delivered to general-purpose PCs. Delivering digital content in a physical medium (like a DVD) represents a combination of the advantages of digital content (e.g., compactness, low manufacturing cost) and the advantages of previous distribution media (e.g., books) in which the content was "bound to" the physical object and hence less easily reproduced. As a result, there is less need to risk the consequences of networked distribution for every work.
Making the content available only on some variety of physical substrate may also have useful consequences for public understanding of the law: If classic movies (or other varieties of high-value content) are known to be available only on disks, then any copy found on the Web is clearly an infringement. This variety of bright-line distinction may be of use in making consumers aware of and more respectful of IP rights.
Conclusion: Some digital information may be distributed more securely using physical substrates rather than by computer networks.
The Digital Millennium Copyright Act of 1998
The Digital Millennium Copyright Act of 1998 (DMCA) amends the Copyright Act, title 17 U.S.C., to legislate new rights in copyrighted works, and limitations on those rights, when copyrighted works are used on the Internet or in other digital, electronic environments.
The anticircumvention regulations adopted by Congress as part of the DMCA need to be clarified to be more technologically sound and more sharply targeted to the problems the regulations were designed to address. The detailed rationale for undertaking such clarification is complex and is presented in Appendix G; a summary appears below.
Certain key terms of the anticircumvention regulations should be defined more precisely. A notable example is the imprecise concept of "an effective technological protection measure." Is a measure that can be circumvented by anyone who has successfully completed a freshman-level college course in computer science an "effective technological protection measure"? What is the threshold for "effective"? Both content owners and potential circumventors need to be able to determine with reasonable effort whether a particular technological mechanism is cov-
ered by the statute and whether a particular act of circumvention is legal. Insofar as these issues are unclear, the statute should be clarified in consultation with appropriate technical experts.23
The encryption research and computer security testing provisions of the DMCA are well intentioned and generally reflect awareness of the importance of research and practice in these fields. But the provisions are also technically unsound in some respects and need to be refined to align better with standard practices in these fields and to allow decisions that are more pragmatic in the context of encryption research and computer security testing.24
Conclusion: More legitimate reasons to circumvent access control systems exist than are currently recognized in the Digital Millennium Copyright Act. For example, a copyright owner might need to circumvent an access control system to investigate whether someone else is hiding infringement by encrypting a copy of that owner's works, or a firm might need to circumvent an access control system to determine whether a computer virus was about to infect its computer system.25
Point of Discussion: Many members of the committee believe in the need to add to the DMCA an exception that would permit circumvention of access controls for "other legitimate purposes." This change would enable judicial discretion in interpreting exceptions to anticircumvention provisions, and would provide needed flexibility in the statute for dealing with legitimate circumvention activities not anticipated by Congress.
The committee's response to and deliberations regarding regulation of what the DMCA calls circumvention technologies reflected the conflicting views evident across the range of stakeholders. Many committee members felt strongly that developing a tool to accomplish any lawful act of circumvention should be lawful, even if the DMCA does not explicitly authorize the development of such a tool. They also believed that a right to develop such tools is implied in the right to engage in lawful circumvention. Others felt strongly that the existing exceptions to anticircumvention rules in the DMCA adequately protect legitimate user interests and regarded as unnecessary any permitting of tool development to pre-
23Other examples of vague and/or technically unsound language in the DMCA appear in Appendix G and in Callas et al. (1999).
24See Appendix G for specifics.
25For additional examples, see Samuelson (1999).
serve fair use. This group feared that allowing development of circumvention tools would inevitably lead to increased infringement.
Some members of the committee felt that people not capable of developing the tools necessary to engage in lawful acts of circumvention should legally be able to acquire the tools from another person, arguing that, in the absence of such an understanding, the explicit right in the DCMA to circumvent would be available only to those with substantial technical skills, thus producing a curious piece of public policy.
Some members of the committee were highly critical of the DCMA's antidevice provisions and suggested that they be repealed; others thought that the provisions should be narrowed in scope. The provisions perceived to be most in need of amendment were sections 1201(a)(2)(ii) and 1201(b)(1)(ii), which outlaw technologies having ''only limited commercially significant purposes or uses other than to circumvent" technical protection measures. However, other committee members felt strongly that section 1201 provides adequately for the use of legitimate devices by responsible persons and that, if a larger class of circumvention devices were legitimized, they would inevitably be distributed widely and used for copyright infringement.
Many members of the committee thought that technologies with noncommercially significant purposes should have the same protection under the DMCA as those with commercially significant purposes. Freeware and shareware, for instance, are examples of technologies that do not generally have "commercially significant purposes" yet may have legitimate, socially desirable uses. These committee members thought that the antidevice provisions would be better phrased in terms of technologies having "no apparently legitimate purpose."
Recommendation: In addition to the currently required Librarian of Congress study of some of the impacts of the Digital Millennium Copyright Act's anticircumvention provisions, broader assessments should be conducted of the impacts of the anticircumvention provisions of the DMCA as a whole. This broader review of the regulations is justified because of their unprecedented character; their breadth; and widespread concerns about their potential for negative impacts on public access to information, on the ability of legitimate users to make non-infringing uses of copyrighted works, on research and development in security technology, and on competition and innovation in the high-technology sector. This review should occur periodically and should include a study of impacts of the anti-device provisions of the DMCA.
As Shapiro and Varian (1998) have pointed out, the appropriate strategy for those in the information business should be to maximize the value of intellectual property, not its protection. Thinking in these terms expands the options available: In addition to the traditional business model of selling digital IP as a product, there are also models that de-emphasize or forgo attempting to control digital information and focus instead on other products or services for which the digital IP is complementary. Additional business models can be developed by asking what forms of value can be derived from the IP that are not so easily reproduced. At the time of this writing, the rapidly evolving Internet has created an effective environment in which to experiment with various business models.
Conclusion: Both technology and business models can serve as effective means for deriving value from digital intellectual property. Technical protection mechanisms can reduce the rate of unauthorized use of IP, but impose their own costs (in production, service, and sometimes customer effort). An appropriate business model can sometimes sharply reduce the need for technical protection, yet provide a way to derive substantial value from IP. Models that can accomplish this objective range from a traditional sales model (low-priced, mass market distribution with convenient purchasing, where the low price and ease of purchase make buying more attractive than copying), to the more radical step of giving away IP and selling a complementary product or service.
Recommendation: Rights holders should give careful consideration to the power that business models offer for dealing with distribution of digital information. The judicious selection of a business model may significantly reduce the need for technical protection or legal protection, thereby lowering development and enforcement costs. But the model must be carefully matched to the product: While the appropriate business model can for some products obviate the need for technical protection, for others (e.g., first-run movies) substantial protection may be necessary (and even the strongest protection mechanisms likely to be available soon may be inadequate).
The Interaction of Technical Protection Services, Business Models, Law, and Public Policy
The community of authors and publishers is characterized by substantial diversity, ranging from those who make their living from the sale of their intellectual property, to those who make their living by keeping it proprietary and using it themselves (i.e., holders of trade secrets), to those who make their careers by giving it away (e.g., most academic researchers), finding reward in the recognition and the indirect benefits that accrue. The differences across all these groups are substantial. There are differences in motivation, in reward structure, and in the consequences of changes in level of IP protection, differences large enough that an attempt to identify the IP solution would be both stifling and counterproductive. Such an approach would also likely focus on the high end of the market, as these products often present the most immediate, compelling, and easily quantified examples of the consequences of IP theft or misuse. But making law or policy by focusing on those examples would be as inappropriate as creating the policy based on the segment of the market that gives away IP and sells auxiliary products or services.
Conclusion: There is great diversity in the kinds of digital intellectual property, business models, legal mechanisms, and technical protection services possible, making a one-size-fits-all solution too rigid. Currently a wide variety of new models and mechanisms are being created, tried out, and in some cases discarded, at a furious pace. This process should be supported and encouraged, to allow all parties to find models and mechanisms well suited to their needs.
Recommendation: Legislators should not contemplate an overhaul of intellectual property laws and public policy at this time, to permit the evolutionary process described above the time to play out.
Moving Beyond the Dilemma: A Call for Research and Improved Data
As the information infrastructure creates many new opportunities and challenges, it is not surprising that research and data collection are needed to support informed decision making. Several such recommendations are made above, including the need to reevaluate the concept of publication; initiate legal, economic, and policy research concerning fair use; and develop an understanding of and data resources about the per-
ceptions and behavior of the general public regarding digital intellectual property. This section identifies additional areas where research and data collection are needed. The committee urges the funders and managers of research programs to place a high priority on the areas of inquiry articulated in this report.
Illegal Commercial Copying
The U.S. copyright industry associations collect a great deal of data and report extensively on illegal commercial copying. Notwithstanding its volume, there are some reasons to question the accuracy of this information. One difficulty derives from the observation that illegal sales and distribution are frequently private acts; consequently, any data necessarily need to be based on extrapolation from a limited sample. There are also methodological issues concerning how the information is generated and reported. Some studies are based on assumptions that lead to high-end estimates, such as the assumption that all illegal copies displace sales at market prices.
A number of committee members concluded that there is a paucity of accurate information and that information reporting needs to be done by a disinterested third party.
Conclusion: The methodology employed by some trade associations in the analysis of data concerning illegal commercial copying produces high-end estimates of losses in gross revenue by these industries. Trade associations would make a more useful contribution to the debate if they revised their methodology so that their estimates better reflect the losses attributable to illegal commercial copying.26 Notwithstanding the methodological deficiencies in the reported information, the volume and cost of illegal commercial copying are substantial.
Multiple, interacting phenomena are at work here. Surely differences exist in the commercial and noncommercial copying spheres, and differences may well exist in behaviors among different demographic groups, different geographic locations, and perhaps even different cultures.27 The phenomena may include such things as how the difficulty of making an illegal copy affects the frequency of copying; the effect on consumer deci-
26See the section "Illegal Commercial Copying" in Chapter 5 for specifics.
27For example, the 14-year-old hackers who download content from the Web without paying should not be lumped with the counterfeiters who contract with a mass-production factory and distribute CDs throughout the world.
sion making of the price and the availability of legitimate copies; the personal sense of the moral or ethical dimensions of the copying involved; the degree of law enforcement or legal scrutiny directed to the behavior; and peer group or other social opprobrium or encouragement. An improved understanding is needed of what these phenomena are and how they operate in the real world, so that they can be targeted for educational efforts and policy-making actions. Reducing the current state of uncertainty about the impact of these various phenomena will be important to future policy makers and entrepreneurs.
Recommendation: Research should be initiated to better assess the social and economic impacts of illegal commercial copying and how they interact with private noncommercial copying for personal use.
The information infrastructure carries both promise and peril for intellectual property; the peril arises from the ability of the technology to make reproduction and transmission of information vastly easier, cheaper, and faster. This, in turn, substantially increases the difficulty of enforcing copyright law. Yet such enforcement is important because of the economic consequences of piracy, the social consequences of laws unenforced, and the belief held by many creators that their works, as expressions of their individuality, ought to be protectable, quite independent of the economic consequences of infringement.
Research on the Economics of Copyright, Use of Patents, and Cyber Law
As a previous CSTB report has observed, "[n]umerous studies have looked at the economic impact of patents, but far fewer such studies have been done on copyright, even though there is currently much legal and policy activity in this area" (CSTB, 1998, p. 49). The committee concurs with this view.
Recommendation: Research should be conducted to characterize the economic impacts of copyright. Such research might consider, among other things, the impact of network effects in information industries and how digital networks are changing transaction costs.28
28Network effects (or, alternatively, positive network externalities) arise when a good is more valuable to a user as more users adopt the same good or compatible ones (Tirole, 1988).
The past decade has seen a substantial de facto broadening of items for which patents can be obtained, including information inventions such as computer programming, information design, and business methods. The long-term effects of this trend are as yet unclear, although the near-term consequences are worrisome.
Conclusion: Because the expansion of patent law to cover information inventions has occurred without any oversight from the legislative branch and takes patent law into uncharted territory, this phenomenon needs to be studied on a systematic basis, empirically and theoretically.
Recommendation: Research should be conducted to ensure that expansion of patent protection for information inventions is aligned with the constitutional intent of promoting the progress of science and the useful arts.
Digital information leads to new kinds of information products and services, which in turn may require legal protection that is difficult to provide through traditional intellectual property law. Digital repositories pose difficult questions about authorship, ownership, and the boundaries among copyright-protected works (e.g., does the information stored on a firm's computer network qualify for protection in the aggregate, or does it consist of many works, some of which qualify while others do not?). Difficulties are also arising with respect to the concept of derivative work, given the mutability of works in digital form and the variety of ways in which digital information can be presented and accessed. Two areas that are particularly likely to generate important legal and policy questions are the status of temporary reproductions and derivative work rights.
• Status of temporary reproductions. Should temporary reproductions of copyrighted works in random access memory be controllable by copyright owners? Much of the debate on this question has occurred as though only two possible positions existed on the issue: that copyright owners are and should be able to control all temporary copies of their works, or that, under present law, they have no right to control any of them. Technologists who think about caching, replication, and the like as ways to build an efficient system view the by-product copies that these activities produce as irrelevant, mechanistic artifacts. Meanwhile, in response to lobbying by particular industry groups about specific temporary copying
issues (e.g., copies made in the course of a transmission by a telecommunications provider), Congress has legislated some specific privileges. Rather than developing the law on temporary copies on a case-by-case, or lobby-by-lobby approach, it would be desirable to develop a taxonomy of temporary copies made in computer systems, and then to assess the economic significance of each category. From this information might emerge some principles about when temporary copies should or should not be regulated by copyright owners, which could then be adapted into a more general-purpose and flexible rule. For example, there might be a workable distinction between "ephemeral" and other temporary copies.29 There is also the question of whether "copies," whether temporary or not, are still the most appropriate basis for copyright; see a discussion of this topic below.
• Derivative work rights. The dynamic and interactive character of digital information raises a host of questions about how the derivative work right of copyright law should be applied. At least one court has thus far taken a fairly narrow view of the derivative work right in the digital environment.30 But filtering, framing, "morphing," real-time language translating, and visualization by other than the rights holder are among the many uses of digital works that raise derivative work rights issues that have yet to be settled.
29A useful principle for distinguishing ephemeral reproduction from ordinary reproductions is that the user can determine the time and circumstances under which an ordinary reproduction is rendered, while this is not the case for an ephemeral reproduction.
30In Lewis Galoob v. Nintendo, the Ninth Circuit Court of Appeals decided that the maker of a "Game Genie" program did not infringe Nintendo's derivative work right by selling a tool with which users could alter certain aspects of the play of Nintendo games. The court held that the Game Genie was not a derivative work because it did not incorporate a protected work or any part thereof in a concrete or permanent form. This ruling suggests that add-on programs will generally not infringe the derivative work right, but many questions remain about how far derivative work rights should extend in the digital environment.
In the view of one writer (Patry, 1994), the Ninth Circuit Court erred because the right to prepare derivative works is not limited to reproduction in copies. Accordingly,"... an unauthorized, unfixed, derivative work will infringe as long as it incorporates a substantial portion of the fixed original work."
In a later case, Micro Star v. Formgen, the Ninth Circuit Court ruled that there was infringement because the derivative work at issue was recorded in permanent form. The court also ruled that there was infringement because Micro Star infringed Formgen's story by creating sequels to that story. The court also considered and rejected Micro Star's argument that it was protected by the fair use defense.
Recommendation: Legal research should be undertaken on the status of temporary reproductions and derivative work rights to inform the process of adapting copyright law to the digital environment, and to assist policy makers and judges in their deliberations.31
As in the case of patents, digital IP and the information infrastructure create new conflicts and raise questions with respect to other information laws and policies.32
The international nature of digital networks contributes to difficulties of applying national laws. Although some work has been done on developing a framework for dealing with conflicts of law, choice of law, and other jurisdictional issues, no international consensus as yet exists on these issues. Thus, legal research should be initiated to help clarify the issues, build consensus, and further the harmonization of intellectual property rules, to promote global information commerce.
Is "Copy" Still the Appropriate Foundational Concept?
The committee suggests above that the notion of copy may not be an appropriate foundation for copyright law in the digital age. Where digital information is concerned, legitimate copies are made so routinely that the act of copying has lost much of its predictive power: So many non-infringing copies are made in using a computer that noting that a copy has been made tells us little about the legitimacy of the behavior. In the digital world, copying is also an essential action, so bound up with the
31Only a few years ago, proxy caching by online service providers and linking from one Web site to another on the World Wide Web were the subjects of considerable debate. Both are now generally thought to be lawful as a matter of U.S. copyright law, a position enabled in part by legal research that has explored the implications of alternative resolutions.
32A new field of law, cyberlaw, is being established, with its own courses and journals.
way computers work that control of copying provides unexpectedly broad powers, considerably beyond those intended by the copyright law.
Recommendation: The committee suggests exploring whether or not the notion of copy is an appropriate foundation for copyright law, and whether a new foundation can be constructed for copyright, based on the goal set forth in the Constitution ("promote the progress of science and the useful arts") and a tactic by which it is achieved, namely, providing incentive to authors and publishers. In this framework, the question would not be whether a copy had been made, but whether a use of a work was consistent with the goal and tactic (i.e., did it contribute to the desired "progress" and was it destructive, when taken alone or aggregated with other similar copies, of an author's incentive?). This concept is similar to fair use but broader in scope, as it requires considering the range of factors by which to measure the impact of the activity on authors, publishers, and others.
The committee recognizes that this undertaking will be both difficult and controversial but suggests, nevertheless, that such an investigation is likely to prove both theoretically revealing and pragmatically useful.
Content Creators and the Digital Environment
The evolution in the information infrastructure presents both potential dangers and opportunities for individual authors of all varieties. A media economy in transition may well produce a realignment of interests, some of which may not be favorable for content creators or publishers. A public policy course of action is needed that maintains a balance between the interests of creators and the interests of those who commercialize intellectual property, with attention given to ensuring that creators will continue to pursue their work.
Because digital works are more malleable than works in other media, new concerns arise about authenticity and integrity. Visual artists have the right to be attributed as creators of their works and have a limited right to protect the integrity of their works, but U.S. copyright law does not expressly protect moral rights of content creators beyond these stipulations.33 Discussion is warranted about what protection might be available to creators of digital works, including the possible role of moral rights.
33The attribution and integrity rights are enumerated in sec. 106A of the Copyright Law.
Conclusion: The digital environment will continue to bring significant and unpredictable consequences for content creators. Further analysis is needed to determine the impact of these consequences and whether any steps should be taken to intervene in the marketplace.
Point of Discussion: Many members of the committee believe that a task force on the status of the author should be established. The goal of such a task force would be to preserve the spirit of the constitutional protection and incentives for authors and inventors. Its mission statement might be as follows: ''The task force shall examine how technological change has affected and is likely to affect the individual creator, recognizing the importance of preserving the economic well-being of creators, balanced with the principle that a democratic society requires broad access to public information." Such a task force would evaluate the viability of mechanisms that facilitate both distribution and control of work (e.g., rights clearance mechanisms) and examine whether issues should be addressed with government action or kept within the framework of private-sector bargaining. The task force would have significant stature with an appropriate level of charter and a limited lifetime (but not less than 2 years); include a cross-section of content creators, rights holders, and other stakeholders; and be financed by public and private funds.
The Process of Formulating Law and Public Policy
The committee has tried, wherever possible, to recommend specific legal and policy actions that will assist in dealing with the digital dilemma. But society is still in the early stages of the ferment brought about by the information infrastructure and still has much to learn about the multiplicity of forces that affect intellectual property. Hence no one can specify with any precision all of the legal or policy actions that will be needed. Where the committee cannot recommend specifics, it has tried to articulate a set of guiding principles that it believes will assist legislators and policy makers in effective formulation and revision of law and policy. This section describes those principles.
Intellectual property and IP protection are primarily conceived as legal constructs, but problems arising in the interaction of IP and the information infrastructure need to be considered in the broader context of other forces as wellmarkets, social norms, and technology (hardware and software). As discussed above, not every problem requires a legisla-
tive solution. Technology, business models, and education can all provide effective mechanisms and means for dealing with problems.
The multiplicity of forces and the new models of content distribution being explored both contribute to the possibility of substitutions. One such substitution is contract law for copyright law: With information products increasingly distributed under license rather than being sold, contract law may begin to substitute for copyright law as the dominant force shaping our information environment. A second substitution has been pointed out by Reidenberg (1998) and Lessig (1999a,b), who note the potential for software to substitute for law. Software can be a form of private regulation, constraining some behaviors just as effectively as legal statutes. The software written to control access to a Web site, for example, can make certain behaviors easy and others nearly impossible.
These sorts of substitutions matter because of what may be gained and what lost in any particular substitution. As the report makes clear earlier, for example, there are both pros and cons in using either contract law or copyright law as a foundation for the information environment. Changing from one to the other should be undertaken with careful consideration of the consequences the shift may bring.
Conclusion: Law and public policy must be crafted to consider all the relevant forces in the digital environment. Initiatives that consider or rely on only one or a subset of the relevant forces are not likely to serve the nation well.
The rapid pace of technological change in the computer industry is the stuff of legend, and it shows no signs of slacking off. This pace is, if anything, increasing. References are made to "Internet time," reflecting the breakneck speed with which business, technology, and social practices change. No appropriately deliberative process has a chance of keeping up. This rapid evolution, particularly in technology, will be an ongoing source of uncertainty and, likely, frustration for policy makers who conceive of and attempt to deal with issues in terms of the extant technology. Such policies are built on shifting sand and run the risk of rapid irrelevance. Even those in the field cannot always cope easily with the pace of change.
Conclusion: Policy makers must conceive of and analyze issues in a manner that is as technology-independent as possible, drafting policies and legislation in a similar fashion. The question to focus on is not so much exactly what device is causing the problem today, as what the underlying issue is. Nor should policy makers base their decisions on the specifics of any particular business model.
Because the information infrastructure makes infringement of IP rights vastly easier, it also makes detection, prevention, and enforcement of laws against private infringement by individuals in their homes far more difficult. As a consequence, individual standards of moral and ethical conduct and individual perceptions of right and wrong become considerably more important in encouraging appropriate behavior. A risk also exists that if IP law is perceived or presented as being so absolute in its prohibitions as to preclude behavior most individuals feel is morally appropriate, then even the more reasonable restrictions in the law may be painted with the same brush and viewed as illegitimate.
Conclusion: Public compliance with intellectual property law requires a high degree of simplicity, clarity, straightforwardness, and comprehensibility for all aspects of copyright law that deal with individual behavior. New or revised intellectual property laws should be drafted accordingly.
Conclusion: The movement toward clarity and specificity in the law must also preserve a sufficient flexibility and adaptability so that the law can accommodate technologies and behaviors that may evolve in the future.
Principles for the Formulation of Law and Public Policy
In addition to the specific guidelines offered above, the committee developed a broader set of principles for policy makers to use in their decision making. The principles are intended to be general and enduring in nature, reflect areas of general agreement, and incorporate the specific guidelines above. Among other things, the principles may serve as a checklist of important issues to consider during the policy decision-making process.
Recommendation: Policy makers should use the principles outlined in Box 6.2 in the formulation of intellectual property law and public policy.
(text box continued on next page)
(text box continued from previous page)
(text box continued on next page)
(text box continued from previous page)
Intellectual property will surely survive the digital age. It is clear, however, that major adaptations will have to take place to ensure sufficient protection for content creators and rights holders, thereby helping to ensure that an extensive and diverse supply of IP is available to the public. Major adaptations will also be needed to ensure that the important public purposes embodied in copyright law, such as public access, are fulfilled in the digital context. Considering the vitality of the participants, the committee is optimistic that workable solutions will be forthcoming in time.
The committee has been cautious about major legislative initiatives because it is early in the evolution of digital intellectual property and much remains unknownboth because of the yet-to-come evolution in the information industries, user communities, and technologies and because of the need for research and data collection to improve knowledge and understanding of the issues. Under such circumstances, major changes in legal regimes and public policy are ill-advised.