Transborder Intellectual Property Issues on the Electronic Frontier Dan L. Burk Visiting Assistant Professor School of Law George Mason University 3401 North Fairfax Drive Arlington, VA 22201-4498 Forthcoming in vol. 5, Stanford Law & Policy Review. Copyright 1994 by Dan L. Burk and the Regents of the Leland Stanford Junior University. This document may be electronically reproduced and disseminated for educational or scholarly purposes. Any commercial use of this document is prohibited without the express written consent of the author and copyright holder. ================================================================ INTRODUCTION The effort to construct a National Information Infrastructure in the United States has stimulated a wave of technical and economic activity that can be expected to extend into the foreseeable future. Because such a "data superhighway" will involve international connectivity, not only the United States, but much of the world will be affected by the NII enterprise. This ferment of new activity, as with every type of human activity, can be expected to generate some disputes that the parties are unable to resolve on their own, and so must look to the law for resolution. For the most part, the legal issues raised by computer networks will be predictable extensions of those raised by print, magnetic, and broadcast media. However, every medium displays its own unique characteristics, requiring some accommodation in the law. In the case of computer networks, such idiosyncracies will include a subtle but important shift in the pattern of incentives related to proprietary information. PROPRIETARY INFORMATION Intellectual property laws exist because of the unusual intangible nature of valuable information. Knowledge and information may be created as a pure intellectual good, or embodied to some degree in a physical form, such as text, pictures, or even as a machine. Such intellectual goods may have considerable value, as evidenced by enormous revenues of the motion picture, recording, or computer software industries. However, creating such valuable goods is not costless; development often requires significant production costs of time and effort. Those who invest in the creation of a valuable commodity generally do so expecting some recompense for their effort. Yet, in the case of intellectual goods, the investor may be denied not only the opportunity to make profits, but even the opportunity to recover production costs. Unlike physical goods, intellectual goods lack the barriers that would allow their investors to prevent free appropriation of the goods by consumers. A manufacturer of a physical good such as shoes can block others from possessing the shoes until paid. By contrast, an idea -- the purest form of intellectual good -- can often be quickly and costlessly acquired by others. The idea will never wear out or require replacement. In addition, the consumer who has the idea can himself become a supplier of the idea, competing with the originator of the idea. No one will pay the creator of the goods if they can get the goods for free. Much like ideas, valuable inventions, songs, or poems can quickly be copied by others, often for very little cost. The physical embodiments of the ideas in these items are nearly as easy to reproduce as a pure idea would be. Because consumers can obtain such intellectual good essentially for free, they are unlikely to pay the developer of the good, and the developer will be unable to recover the development costs for the good. Developers who realize this will be reluctant to invest in developing goods that everyone can have for free, and so there may be a lack of intellectual goods. Intellectual property laws supply the barriers to consumption that intellectual goods lack. Patent, copyright, and trade secret laws create intellectual property rights, imposing legal and financial penalties for appropriating an intellectual good without the owner's permission. Because the developer of an intellectual good can use the intellectual property right to exclude consumers, the developer can expect a return on his investment, and so the production of valuable and useful intellectual goods is encouraged. These legal barriers will become increasingly important in the context of global computer networks such as the Internet. THE INTERNET The Clinton administration has called for construction of a publicly available "information superhighway," but the first generation of such a network already exists in the form of the Internet. Information flow on the Internet resembles that in a river, highway, or circulatory system: local networks funnel information traffic into larger regional networks, which in turn are connected to high capacity "backbone" linkages. This network of computers, its electronic traffic enabled and directed by host computers at each interconnected site, links literally millions of users in dozens of countries all over the world. The electronic environment of the Internet magnifies and intensifies associated with information development incentives. The Internet, as an"information superhighway" facilitates the exchange of intellectual goods embodied in easily reproducible form: electronically encoded text, data, pictures, music, and software. Once the network connections are in place, the transfer of data costs practically nothing. Electronic archives on the Internet allow thousands of individuals to access information at a marginal cost close to zero. Materials in electronic format can be reproduced and distributed just as cheaply. This is an enormous benefit to the transfer of information among individuals, prompting the present administration's calls for increased Internet access and capacity. However, these benefits also have a dark side arising from the incentives associated with the creation of valuable information: as described above, there is little incentive to invest in developing information where no return on the investment is likely. Where everything is free, little will be offered. These problems of reproduction and dissemination, even though magnified by the capabilities of the Internet, seem to be familiar in kind, if not in scope: they are the type of problems that intellectual property laws are designed to remedy. Thus, the need for intellectual property rights in this networked environment seems manifest. Yet, these laws as presently formulated are geared to a world in which barriers of distance and political jurisdiction are the norm. This is not the world of the Internet, where computer-enabled features such as remote log-on and telnet allow users to control computers, radio telescopes, electron microscopes, and other scientific instruments from a distance. Computers on the Internet can be linked together to simultaneously process different parts of complex problems, creating in effect a giant supercomputer. These features of the network make distance and political borders invisible to users of the electronic universe they have dubbed "cyberspace." This ability to reach out across vast distances to connect remote users and equipment is novel to the global computer network. Modern telecommunications have for some time allowed broadcast and transport of information over wide areas, but only the capabilities of a computer network allow routine control of activity across national boundaries. This creates a new challenge for intellectual property regimes, the problem of remote access. The law in general, and the law of intellectual property in particular, has failed to take stock of this phenomenon. Law is in general territorial, extending no farther than the borders of the jurisdiction whose government has enacted that law. Cyberspace extends beyond those borders, however, and this creates new incentives regarding information transfer. (1) For example, one can readily envision computer sites linked to the Internet, offering information services, but lying beyond the political territory of the United States. Such offshore information sites would lie beyond the reach of U.S. intellectual property laws, but not beyond the reach of information product consumers in the United States. Consequently, offshore havens could offer information services that would violate U.S. laws if the sites were in the U.S. Such illicit services can be expected to arise as a simple function of economics; where the cost of access to the offshore information is lower than the cost of obtaining a license from the legitimate owner of the information, users will naturally choose the cheaper alternative. It seems likely that the cost of using the offshore version will be extremely low because the incremental cost for data exchange over computer networks is very small. In order to make a licensed version of his information competitive with that available offshore, an information owner would have to price his license lower than the cost of Internet access. However, at a price that low, the information owner is unlikely to recover his development costs. As with the problems of reproduction and distribution, this result of remote access may deter information developers from investing in generating the information in the first place. Thus, the Internet gives rise to two distinct sets of intellectual property problems, stemming from the nature of the network: not merely an electronic communication linkage, but a collection of digital information processors. First, the Internet allows directed and selective transfer, dissemination, and retrieval of information. Second, the remote logon and telnet features of the Internet allow users in one geographic location to access and control devices in far distant geographic locations. Each of these sets of features will challenge the current regime of patent, copyright, and trade secret laws in distinctive ways. PATENTS Perhaps the most important set of laws protecting intellectual property are the statutes authorizing the set of rights called patents. A patent allows the holder to exclude others from making, using, or selling the patented invention for a period of seventeen years. (2) Under U.S. law, patents may be granted for machines, processes, and compositions of matter. (3) Significantly for the computer industry, "processes" under the patent law may include computer software -- software is after all a set of sequential instructions to a machine. Thus, the United States Supreme Court has cautioned that computer algorithms in the abstract may not be patented, but may be claimed as part of the function of a machine. (4) Unlike other forms of intellectual property, patent rights do not arise spontaneously with the creation of protectable subject matter. Rather, patents are granted after examination of an application by a government agency, the Patent and Trademark Office. Applicants must state specific claims as to the invention that they wish to patent; after they are approved, the terms of the claims will determine the scope of the patent, much as a deed sets out the metes and bounds of real property. The examination process assesses whether the claimed invention meets statutory criteria of novelty, usefulness, and non- obviousness. (5) In particular, the examination process compares the claimed invention to "prior art," that is, to previously known inventions in the same field, to determine that the new invention is a significant advance over the old. United States patent law is territorial; it applies only to making, use, or sale of the claimed invention within the United States, its territories, and possessions. (6) According to the United States Supreme Court, this limitation prevents the patent law from offending the sovereignty of other nations. (7) However, the limitation also engenders a version of the remote access problem suggested above: Internet users located in the United States could access offshore computer sites that employ patented U.S. software or hardware. (8) If data were retrieved from or processed at such sites, the use of the patented invention could be obtained, but only unpatented information would cross into U.S. territory. Conversely, Internet users outside the United States, using computer hardware or software that would infringe U.S. patents, might use such equipment to access Internet resources within the United States. Patent law in its present state is indeterminate as to whether such transborder activity constitutes an infringing "use." Certainly the activity has some connection with U.S territory, but that connection may be too tenuous to trigger enforcement of the patent. Some older patent cases dealing with "extended instrumentalities" -- such as a navigational broadcast system using transmitters outside the United States -- indicate that a transborder invention will be viewed as constructively "within" the United States if primarily controlled from U.S. territory. (9) However, it is unclear how applicable these older cases to may be to the case of global computer networks, particularly if it is uncertain where "control" is being exercised on the network. Questions of extraterritorial infringement have also to some extent been addressed though customs statutes prohibiting importation of products that infringe a U.S. patent, or of products produced abroad via processes that would infringe a U.S. patent if the process were practiced in U.S. territory. (10) Here again, however, it is unclear that the statutes would apply to extraterritorial infringement via computer network. Data processed by an offshore site and transmitted into the United States might in some sense be considered "imported" and, as the result of software patented in the United States, is perhaps the product of an infringing foreign process. It is doubtful, however, that Congress in fact considered such applications of these statutes when enacting them, and applying the statutes to computer networks may stretch the law beyond its proper bounds. The conundrums posed by remote access may affect not only the enforcement of patents, but also the procurement of patents.Under United States patent law, the events that establish priority of invention may not be established by reference to foreign activity. (11) Similarly, although the patent law recognizes printed material from anywhere as "prior art," tangible items are recognized as prior art only if they were previously known, used, or invented "in this country," that is, in the United States. (12) Additionally, placing an invention "in public use" more than one year prior to applying for a patent on it renders the invention unpatentable, because the invention effectively becomes its own prior art. (13) This rule of patentability, like those regarding priority, applies only to activity "in this country." It is quite easy to envision situations where a U.S. researcher, working in the United States, could via the Internet use equipment situated in another country for experiments leading to a patentable invention. It is unclear, however, whether these events could be used to establish priority to claim a patent right in the invention. Conversely, a U.S. patent applicant might be able to obtain a patent to an invention that was already well known and routinely available via Internet at an offshore site; it is not clear that the offshore invention would constitute prior art if never physically used "in this country." Once again, older patent cases dealing with "extended instrumentalities -- such as a remotely controlled satellite in earth orbit -- may be considered constructively within the United States for purposes of patent procurement, but the applicability of these cases to the Internet is uncertain. Finally, the Internet's electronic dissemination capabilities may also impact patent procurement. In the United States, complete disclosure of an invention more than one year before filing for a patent renders the invention ineligible for the patent. (14) At present, it is unclear what might constitute an Internet "publication" of information sufficient to prevent patenting. Past cases have focused on the aspect of public accessibility; information contained in an uncataloged thesis was not publicly accessible, and so not a disclosure, whereas information in a cataloged thesis was publicly accessible, and so barred a patent. On the Internet, many different types of electronic archives are available, but they are at present extremely user unfriendly, and information is difficult to find unless one knows before where to look. However, Internet-wide electronic searching is becoming increasingly feasible through software such as ARCHIE, VERONICA, MOSAIC, or WAIS. For a publicly accessible national information infrastructure to be fully implemented, this trend toward increased accessibility will have to continue. However, increased public accessibility of information may have the collateral effect of barring many inventions from patentability. COPYRIGHT Copyright comprises a second distinctive set of intellectual property protections with important for transborder computer networks. As its name implies, copyright secures the right to reproduce a protected work. (15) Copyright does not protect ideas, but rather the particular expression in which ideas are embodied. Thus, copyright is likely to affect most of the types of information exchanged over the Internet: the subject matter protected under copyright law includes music, pictures, text, and graphic designs. Significantly, copyright offers an alternative to patent for the protection of computer program code. Copyright also protects the order and arrangement of databases if the arrangement of the information meets a certain threshold of creativity and originality. (16) Because the essence of copyright is the ability to exclude others from reproducing the protected work, the anomalous effects of an international networked environment will be somewhat less pronounced than in the case of patent protection. Retrieving copyrighted information from an offshore source will almost certainly involve infringement; for the information to be useful, an infringing copy will at some point materialize within the jurisdiction, as transferred computer code, a screen display, or in some other form. In addition, unlike patent law, copyright law is subject to international treaties such as the Berne Convention that impose recognition and reciprocity of copyrights among the signatory nations. There is still some variation between nations as to specific copyright provisions, but the international treaties guarantee certain minimum standards of protection. Yet it is still apparent that territorial anomalies may be exploited to evade copyright in some instances. (17) Not all nations are signatory to the international copyright treaties, and many developing nations are notably absent. For such nations, maintaining minimal copyright recognition or lax copyright enforcement may seem viable strategies to capture international information business. For example, some valuable databases comprise information that is not copyrightable because it lies in the public domain; such databases might include those compiled by Mead Data and by West Publishing for on-line legal research. The value of these databases lies in their order, format, and arrangement, all of which is copyrightable subject matter, even if the substance of the databases is not. It is not difficult to conceive of an off-shore service offering an identical database in a nation that does not recognize copyright protection for the arrangement of the data. Database users in the United States could access the off-shore database and retrieve the information without violating copyright law: the substantive data brought into the United States is not the subject of a copyright, and the database -- which is the subject of copyright -- lies outside U.S. jurisdiction. Analysis of copyright law also diverges from that of patent law where, in contrast to the territorial focus of patent law, copyright statutes tend to follow principles of nationality. Because of the Berne Treaty, much of U.S. protection of a copyrighted work is dependent the nationality of the author rather than on the location in which it was created. For example, section 104 of the copyright act provides that unpublished works are subject to U.S. copyright protection without regard to the nationality of the author, whereas published works are protected only of certain nationality criteria are met. These criteria require that the work be produced by a U.S. national, a foreign national from a country with which the United States has signed a copyright treaty, or a "stateless person." Similarly, section 101 provides that an unpublished work is a "Berne work" for purposes of the copyright laws if one or more of the authors is a national of a Berne treaty state. Section 101 also provides the United States is considered the country of origin of a work of all the authors are nationals of the United States. However, the copyright law does partake of some aspects of territoriality. Under section 101, the United States is considered the country of origin of a work if it is published "in the United States." Similarly, in order for a published work to qualify as a Berne work under section 101, the work must be published in a Berne treaty nation. These distinctions are important because at present, U.S. copyright laws impose different obligations upon works originating in the United States. For works originating in the United States, registration and deposit of the work with the copyright office is required in order to obtain statutory damages and attorney's fees in an infringement action. As with the patent law, the territorial aspects of copyright law become aberrant in the face of the Internet's remote access features. The capabilities of the Internet allow an author present within one jurisdiction to create a copyrightable work, such as a computer program or data file, within another jurisdiction, and to publish the work in yet another jurisdiction. The assumption built into the statutes, that the author and her creation will be present at the same physical location is no longer valid, and leaves the status of such a work uncertain. The assumptions inherent in the copyright law with regard to authorships seem absent with regard to publication, and with good reason. The author and the publisher of a work may be physically distant even under established technologies. However, because of its electronic dissemination features, the Internet introduces a different set of anomalies with regard to publication. The status of "publications" is critical in copyright; among other reasons, the period of the copyright for some works begins to run as of the date of publication, and the ability of libraries and archives to reproduce a work depend upon its publication status. (18) Much as in the case of public disclosure under patent law, copyright publication hinges upon public availability. However, copyright has long recognized "limited publication," by which a work is disseminated to only a select group of individuals for a restricted purpose. The various electronic fora on the Internet, such as bulletin boards, listserv subscriptions, and USENET newsgroups vary widely in their purposes and degree of public accessibility. Thus, some dissemination on the network will constitute copyright "publication," but by no means all, and the details of what constitutes publication on the network may take some time to clarify. TRADE SECRECY An additional form of intellectual property protection exists under the rubric of trade secrecy. Trade secrecy protects any type of information that is not generally known to the competitors of the owner, and which by virtue of its proprietary nature gives the owner a business advantage. (19) The trade secret need not be an actual secret; the owner need only take reasonable steps to keep the information confidential. If the information is made publicly available, whether by the owners purposeful disclosure, inadvertent disclosure, or independent discovery by another, the information passes into the public domain and trade secret protection is lost. In some instances, trade secrecy may offer an alternative to patent or copyright protection of computer software, as for example in the case of a computer operating system whose source code is not published or easily accessible for reverse engineering. The primary context in which trade secrecy is operative is between competing businesses, where it offers legal protection against industrial espionage and against disclosure of proprietary information by a former employee to the competitor of a the former employer. It prevents competitors from "reaping where they have not sown"; a business may not free ride off the research investment of its competitor. Consequently, reverse engineering or independent discovery of a competitor's trade secrets are not barred by trade secrecy, only practices that appear to entail theft of information are prohibited. In a networked environment, the characterization of trade secret infringement as "information theft" means that this type of intellectual property protection is inextricably tied to the broader question of computer crime. Indeed, in many jurisdictions, theft of trade secrets is penalized by criminal sanctions. Unfortunately, deterrence of computer crime is an obdurate problem, especially in the international context. Many nations have no law of trade secrecy, or pay little attention to enforcing the laws they have. Additionally, even jurisdictions with sophisticated trade secrecy laws may not yet have come to grips with the concept that electronic property may be stolen; these gaps in enforcement, coupled with the Internet's remote access capability, may effectively create computer piracy havens. For example, computer hackers based in German territory were able to break into U.S. computer sites with relative impunity, because German law did not prohibit taking of U.S. data. (20) These problems are by no means limited to illicit movement of information across international borders. Much the same situation exists with regard to transborder flow within the United States. Trade secret law is state law, not federal law, and so varies between states. Many jurisdictions have enacted computer crime statutes that penalize the most blatant type of trade secret theft: breaking into a computer from outside in order to remove valuable information. However, trade secrecy violations tend to emerge from inside, rather than outside businesses, as in the celebrated case of the "nuPrometheans": certain disgruntled engineers at Apple Computer Corporation, calling themselves the "nuPrometheans" took it upon themselves to make the Apple MacIntosh computer's proprietary operating software freely available to a variety of parties. Apple was fortunate to be able to retrieve all the distributed copies before they became publicly available and trade secrecy in the software lost. In such instances, the applicability of the state computer crime statute will depend on how "authorization" has been defined -- the employee may have authorized access for an unauthorized purpose. In addition, the presence of a "property" interest in the stolen information will depend upon the particular state's definition of trade secrecy. This lack of uniformity generates enormous difficulty in determining not only what constitutes protected information, but also the type and the level of protection the information will receive. For example, the definition of trade secrecy in the Restatement of Torts, a definition adopted by many states, requires that the secret be actively used in the holder's business; the definition found in the Uniform Trade Secrets Act, a definition adopted by many other states, imposes no such requirements. (21) Some federal statutes, such as the National Stolen Property Act, penalize theft and transport of property across state lines; it might be expected that these laws could be applied to electronic theft as well as physical theft. Such application of these laws appears problematic, however. (22) The Stolen Property Act, for example, criminalizes the transport of stolen "goods" in interstate commerce, but it is unclear whether Congress contemplated electronic transmission of trade secrets within the meaning of "goods" under the statute. The ephemeral nature of such trade secrets has been particularly vexing in this regard. Where trade secrets such as a secret chemical formula are stolen in documentary form, courts have not hesitated to apply the Stolen Property Act. (23) Similarly, at least one court has applied the Act in a case where proprietary information was removed from a computer via unauthorized access, and the information was posted on a computer bulletin board in another state. (24) More recently however, when a former employee stole computer source code, the software was held not to be a "good" under the Act, because of its intangible nature. (25) Consequently, no uniform protection of trade secrets is presently available even within the United States. INDICATIONS Traditionally under international law, nations have asserted sovereignty based upon the territory that they could control. The advance of electronic communications has begun to change this basis for sovereignty, and the Internet has accelerated the change. The legal conundrums reviewed here suggest that, rather than sovereignty based on territory, sovereignty based on information flow or economic spheres of influence will be the norm in cyberspace. This shift will require a reevaluation of present legal doctrines. Clearly, the present U.S. intellectual property provisions, when stretched to fit this new jurisdictional terrain, display significant gaps in their coverage. One approach might be to acknowledge the inexorable erosion of territory- based sovereignty, and draft statutes that operate without regard to political borders. For example, actions rather than location might form the basis for asserting jurisdiction. A statute based on this approach would effectively extend the force of United States law to electronic actions that affect U.S. interests, regardless of the physical situs of the actor. Some precedent for this exists in international law, where the doctrines of protective jurisdiction and passive personality jurisdiction serve as the bases for jurisdiction over foreign nationals whose actions adversely affect the interests of a nation or its citizens. (26) However, such bases for asserting jurisdiction have been controversial, and are likely to be even more so in the uncertain arena of computer networking. Any statute that extends United States jurisdiction without reference to political boundaries will be seen as a violation of those boundaries, and as an affront to other nation's territorial sovereignty. Electronic data exchange may be eroding political borders, but it has not yet abolished them. It is unlikely that the world is ready to accept the inevitable, and statutes based on non-territorial sovereignty are likely premature. The alternative is, for the present, to fix the existing territorial statutes piecemeal as new problems are identified. Courts can handle much of the load; where uncertainty as to the application of a legal doctrine rests upon factual questions, judicial interpretation should serve to accommodate the idiosyncracies of cyberspace. For example, case by case delineation of the electronic actions that constitute "publications" for copyright purposes, or "disclosure" for patent purposes, is a task well suited to the judiciary. In other instances, particularly those involving remote access, truly novel problems may be presented, and these will require broader policy decisions. In such cases, statutes will likely require amendment to plug their holes. Thus, Congressional reconsideration of terms such as "use" under the patent statutes, or "goods" under the National Stolen Property Act would seem to be in order. Some new enactments would also seem to be in order; for example, it may be advisable for Congress to enact the Uniform Trade Secrets Act for application to electronic interstate commerce. Such piecemeal provisions should be sufficient to shore up the legal structure until a more permanent solution is possible. ================================================================ NOTES 1) See Nicholas P. Miller & Carol S. Blumenthal, Intellectual Property Issues in Toward a Law of Global Communications Networks 227, 228 (Anne W. Branscomb ed., 1986). 2) 35 U.S.C. sec. 271(a) (1988). 3) 35 U.S.C. sec. 101 (1988). 4) See Diamond v. Diehr, 450 U.S. 175 (1985). 5) See 35 U.S.C. secs. 101-03 (1988). 6) 35 U.S.C. sec. 100(c) (1988). 7) Deepsouth Packing Co. v. Laitram Corp., 406 U.S. 518 (1972). 8) See Dan L. Burk, Patents in Cyberspace: Territoriality and Infringement on Global Computer Networks, 68 Tulane L. Rev. 1 (1993). 9) See Decca Ltd. v. United States, 544 F.2d 1070 (Ct. Cl. 1976). 10) 35 U.S.C. sec. 271(g) (1988). 11) 35 U.S.C. sec. 104 (1988). 12) 35 U.S.C. sec. 102(a) (1988). 13) 35 U.S.C. sec. 102(b) (1988). 14) Id. 15) 17 U.S.C. sec. 103 (1988). 16) See Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991). 17) See Ithiel de Sola Pool & Richard Jay Solomon, Intellectual Property and Transborder Data Flows, 16 Stan. J. Int'l L. 113, 124-25 (1980). 18) See 1 Raymond Nimmer, Nimmer on Copyright sec. 4.01[A] (1989). 19) See Restatement of Torts sec. 757, cmt. a (1939). 20) See Anne Wells Branscomb, Jurisdictional Quandaries for Global Computer Networks in Global Networks: Computers and International Communication 84, 92 (Linda M. Harasim ed., 1993). 21) See Uniform Trade Secrets Act, commentary to sec. 1, 14 U.L.A. 433 (1985). 22) See Seth Greenstein, Is Software "Goods"? Intangibility Revisited Under the UCC and the National Stolen Property Act, 8 Comp. Lawyer 30 (1991). 23) See United States v. Greenwald, 479 F.2d 320 (6th Cir. 1973). 24) United States v. Riggs, 739 F. Supp 414 (N.D. Ill. 1990). 25) United States v. Brown, 925 F.2d 1301 (10th Cir. 1991). 26) See Dan L. Burk, Application of United States Patent Law to Commercial Activity in Outer Space, 6 Santa Clara Comp. & High Tech. L.J. 295, 316-17 (1991). .