DRM and Privacy

Julie E. Cohen* © 2003, Julie E. Cohen. Copies of this article may be made and distributed for educational use, provided that: (i) copies are distributed at or below cost; (ii) the author and the Berkeley Technology Law Journal are identified; and (iii) proper notice of copyright is affixed.

I. Introduction

The future of privacy is increasingly linked to the future of copyright enforcement. In their push to control the proliferation of unauthorized copies, and to maximize profit from information goods distributed over the Internet, copyright owners and their technology partners are designing digital rights management (DRM) technologies that will allow more perfect control over access to and use of digital files. The same capabilities that enable more perfect control also implicate the privacy interests of users of information goods. Although DRM technologies vary considerably, at the most general level they represent an effort to reshape the practices and spaces of intellectual consumption. They also create the potential for vastly increased collection of information about individuals’ intellectual habits and preferences. Quite apart from the questions of intellectual property policy that surround DRM technologies, therefore, the proper balance between DRM and user privacy is an important question in its own right.

Interrogating the relationship between copyright enforcement and privacy raises deeper questions about the nature of privacy and what counts, or ought to count, as privacy invasion in the age of networked digital technologies. This Article begins, in Part II, by identifying the different types of privacy interests that individuals enjoy in their intellectual activities and exploring the different ways in which DRM technologies threaten these interests. Part III considers the appropriate scope of legal protection for privacy in the context of DRM, and argues that both the common law of privacy and an expanded conception of consumer protection law have roles to play in protecting the privacy of information users. Finally, Part IV considers whether DRM technologies and standards processes themselves might be harnessed to protect privacy.

II. Privacy Interests in Intellectual Consumption

DRM technologies operate at the intersection of two complex and powerful constellations of privacy values. They target a set of behaviors, which I will label intellectual consumption, that often (though not always) take place within private spaces. These behaviors, in turn, concern an activity – intellectual exploration – that is widely regarded as quintessentially private.

Within most non-primitive societies, tradition and social practice reserve certain types of “private space” to the individual or the family. Chief among these is the home, which is conceived as a place of retreat from the eyes of the outside world. Some privacy skeptics argue that rules about entitlements to privacy within certain spaces overlap substantially with property-based entitlements to control access to private homes or offices.[1] Yet the correspondence between ownership and spatial privacy is imperfect. Not every invasion of a residential property interest is an invasion of privacy; for example, most people do not think that a nuisance, such as excessive noise or noxious fumes, is also a privacy invasion.[2] And individuals can have privacy expectations in spaces that they do not own or rent, such as public restrooms, dressing rooms, and telephone booths.[3] Acknowledgment of these expectations suggests a fairly broad consensus that the interests protected by “privacy” and “property” are different. Rules and traditions about freedom within private spaces concern not only property interests, but also guarantees of breathing space for individual behavior. One may, for example, walk around nude inside one’s own home, even though one is not free to do so in public. In just the same way, spatial privacy allows for metaphorical nudity; behind closed doors, one may shed the situational personae that one adopts with one’s co-workers, neighbors, fellow commuters, or social acquaintances, and become at once more transparent and more complex than any of those personae allows.[4]

Traditions of privacy in intellectual exploration are more recent, but equally robust. Within Western societies, a central tenet of post-Enlightenment thought is the inviolability of each individual’s rights over her own person.[5] These rights include not only rights of bodily integrity and other corporeal rights, but also rights over one’s own thoughts and personality. Surveillance and compelled disclosure of intellectual consumption threaten these rights in subtle but powerful ways. Although a person cannot be prohibited from thinking as she chooses, persistent, fine-grained observation subtly shapes behavior, expression, and ultimately identity.[6] The inexorable pressure toward conformity generated by exposure, and by loss of control over uses of the gathered information, does not so much invade right of self-determination as it coopts them. Additionally, surveillance and exposure devalue the fundamental dignity of persons by reducing the exposed individuals to the sum of their “profiles.”[7] For these reasons, in other circumstances where records of intellectual consumption are routinely generated – libraries, video rental memberships, and cable subscriptions – society has adopted legal measures to protect these records against disclosure.[8] Just as spatial privacy guarantees breathing space for behavior, privacy rights in the information generated by intellectual exploration guarantee breathing space for thought.

DRM technologies are poised to affect both the spatial and the informational dimensions of privacy. Both by directly constraining private behaviors related to intellectual consumption and by enabling creation of detailed and permanent records of such consumption, these technologies have the potential to change dramatically the way people experience intellectual goods. Whether they will do so in a way that undermines either set of privacy values is an important question. To answer it, we must consider each of the general types of functions that a DRM technology might perform.

A. Constraint

Some DRM technologies are designed to set and automatically enforce limits on user behavior. For example, a music delivery format might prevent copying, including copying for “space-shifting” purposes, or might restrict the types of devices that can be used for playback.[9] The “content scrambling system” (CSS) algorithm used on DVDs does both of these things, and also implements a “region coding” compatibility system designed to ensure that DVDs intended for use in one geographic region (e.g., North America) cannot be played on equipment sold elsewhere.[10]

Technologies that force changes in user behavior narrow the zone of freedom traditionally enjoyed for activities in private spaces and activities relating to intellectual consumption. In so doing, they decrease the level of autonomy that users enjoy with respect to the terms of use and enjoyment of intellectual goods. Does this constriction also amount to an invasion (or, more neutrally, a lessening) of privacy? That depends on how privacy and its absence are defined.

It is hard to argue that a copy-protection device “intrudes on seclusion” in the precise manner contemplated by the Prosserian tort of that name. The tort theory of spatial privacy envisions “seclusion” as physical isolation from observation. The sort of intrusion cognizable as privacy invasion generally involves direct human agency and at least the possibility of a human observer. Technologies of direct constraint, in contrast, operate automatically and without recourse to an external controller. But to say that, therefore, these technologies cannot “intrude” begs the question whether standards devised by courts to remedy invasions of private space in the predigital age should be the touchstone for assessing diminutions of spatial privacy in the digital age. A less precedent-bound approach to defining privacy and its absence/loss might view matters differently.

More abstractly, many philosophers conceive of “privacy” as a condition of inaccessibility or limited accessibility to the rest of the world.[11] Invasions of privacy involve rendering the individual more accessible to others in some way. Technologies of direct constraint do not map especially well to this theory, either. Copy-control restrictions and similar constraints do not render individuals who purchase restricted works more accessible to others in any particularized way; they simply carry out their assigned tasks. If I buy a copy-protected music CD and play it in my living room, I and my living room are no more accessible to the copyright owners of the various musical works and sound recordings than the day before I made my purchase.

Conceptualizing loss of privacy in terms of either intrusion or particularized accessibility, however, misses an important aspect of the dynamic established by DRM technologies of direct constraint. There are several possible ways to respond to the problem of policing under conditions of limited accessibility. One is to develop DRM technologies that enable surveillance; those technologies are discussed below. Another – the strategy of direct constraint considered here – is to restrict the range of permitted behaviors in a way that is known ex ante, thereby eliminating any need for intrusive monitoring.[12] This strategy subverts the privacy-accessibility dynamic. I and my living room may be no more accessible to the copyright owners of the copy-protected music CD than before I bought it, but that does not matter; the feasible uses of the CD are known, and so the question of particularized accessibility to me is moot. Yet it is hard to see the result as non-invasive; if anything, it is more efficiently invasive than a surveillance strategy would be.

Focusing narrowly on “intrusion” or “accessibility” also ignores the complex intersectionality of the privacy concerns implicated by DRM technologies; it emphasizes spatial privacy but excludes consideration of intellectual privacy values. Some philosophers argue that “privacy” is not only a condition of (relative) inaccessibility, but also, in part, a condition of autonomy with respect to important personal decisions.[13] Although the usual examples relate to rights in one’s own person (e.g., decisions about reproduction, or about intimate relationships), one might extend this argument to encompass rights in one’s own mind. Particularly where intellectual exploration is involved, privacy concerns guarantees of “breathing space” for individual choice. Technologies of direct constraint shape individual practices of intellectual consumption in ways that shift the locus of choice about those practices away from the individual.

One might object that this argument makes every product design problem a privacy problem, and that this result does not square with the realities of the competitive marketplace. According to this view, DRM technologies of constraint, like any other new consumer product feature, simply create for users new realities around which to exercise (fewer remaining) choices. This, though, presumes that “product design” results from a confluence of neutral/technical factors exogenous to social policy. My point is exactly the opposite. Product design reflects social as well as “technical” values – or perhaps more precisely, technical considerations cannot help but reflect social ones.[14] For an example, one need look no farther than DRM technologies themselves; design for maximum constraint reflects commercial and (anti)competitive objectives. There is no reason to say that privacy does not belong in the calculus of factors that inform and constrain design; quite the contrary is true. If spatial privacy and intellectual privacy are important human values, then product design is a privacy issue, and rightly so. It follows that sometimes privacy values will receive only partial accommodation; one cannot say that privacy is the only relevant design consideration. But one can articulate as an explicit norm of the design process the goal that delivery technologies for information goods should be designed to minimize privacy-invasive constraints. As I discuss in great detail in Part IV, injecting this norm into the DRM design process might produce DRM technologies that look substantially different.

B. Monitoring

Other DRM technologies are designed to report back to the information provider on the activities of individual users. Such reporting may occur in conjunction with a pay-per-use arrangement for access to the work, or it may occur independently of payment terms. For example, monitoring functionality might be designed to collect data about use of web browser software, or about the types of documents created using word processing or spreadsheet software.[15] Monitoring also can be used to determine information about related products, such as the presence of non-copy-protected mp3 files on the user’s hard drive or the other computer programs a user is running in conjunction with a licensed program.[16]

DRM technologies that monitor user behavior create records of intellectual consumption. Indirectly, then, they create records of intellectual exploration, one of the most personal and private of activities. They also create records of behavior within private spaces, spaces within which one might reasonably expect that one’s behavior is not subject to observation. Intuitively, the case for regarding these technologies as privacy-invasive seems clearer than it did for technologies of direct constraint.

Compared with DRM technologies of direct constraint, the monitoring capabilities of DRM technologies fall more straightforwardly within conventional definitions of privacy invasion. Gathering information about intellectual consumption renders intellectual preferences accessible, both to the information provider and to third parties that might purchase it or, in the case of the government, compel its production. Much of this record-keeping activity is conducted automatically, without the direct involvement of a human observer or controller, but the fact of automation does not neutralize the threat to privacy interests. The relevant question, instead, is whether information about intellectual consumption is gathered and stored in a form potentially accessible to others. Whenever those conditions hold, privacy interests have been compromised. For the same reasons already discussed, one might argue, as well, that DRM monitoring technologies invade privacy interests grounded in decisional autonomy considerations, by rupturing the breathing space required for intellectual exploration.

DRM monitoring technologies also can have second-order privacy effects. Specifically, data gathered through monitoring can later be used to generate detailed profiles of users’ revealed intellectual preferences. The information provider can use the resulting profiles to market additional information goods to users, or can sell it to third parties who may use it for a wide variety of other purposes.[17] DRM monitoring technologies do not uniquely enable profiling, or even intellectual profiling; without any information about usage patterns, an information provider can construct a reasonably detailed profile of intellectual preferences and subject matter interests based solely on the information generated by initial purchase records. Nonetheless, the use of data gathered via DRM monitoring to “enhance” existing profiles renders those profiles more comprehensive, and thus potentially more invasive from the user’s perspective for all of the reasons already discussed.

C. Self-Help

Direct restriction protocols can be designed to encode penalties as well as disabilities. For example, a DRM system could be designed to disable access to a work upon detecting an attempt at unauthorized use. Such “self-help” technologies – so named because they are designed to obviate recourse to legal enforcement procedures – may be directed and controlled externally upon detection of the prohibited activity. This type of functionality must be implemented in tandem with some sort of monitoring functionality. Self-help technologies also can operate automatically upon internal detection of a triggering activity, without communication with any external system or controller. The extent to which either type of self-help functionality should be permissible as a matter of contract law has been the subject of an ongoing dispute,[18] but there appear to be no technical barriers to their implementation.

DRM self-help technologies present a special case of the constraint problem, and potentially a special case of the monitoring problem as well. For all of the reasons already discussed, I believe that it is analytically sound to conclude that both types of technologies significantly diminish privacy in intellectual consumption. There remains the question whether the inclusion of self-help functionality adds anything distinct to the privacy dynamic.

The punitive quality of self-help implicates privacy interests in one way that technologies of direct constraint do not. The identification of a particular consumer as a target for self-help measures entails loss of the (partial) anonymity that individual formerly enjoyed as one among many customers.[19] Here again, DRM technologies give the dynamics of enforcement a slightly different spin. Enforcement, like constraint and monitoring, can be activated without direct human agency; thus, it is conceivable that no human need know the specific identities of those singled out. Here again, though, conceptualizing loss of privacy in terms of human “attention” misses the distinctive sense in which the phenomenon of attention operates in the digital age. Attention and anonymity, or at least fungibility, may coexist. One can remain an anonymous customer and yet be singled out by a process of automated decisionmaking for consequences that one would not choose.[20] Whether a human or a computer directed the decision, one’s ebooks and mp3 files no longer “work,” and no longer work as a result of actions taken privately. From the individual user’s perspective, the consequences are the same regardless of whether a human or a computer made the final call to activate self-help measures.

In addition to the loss of particularized privacy for targets of self-help measures, the employment of DRM self-help technologies raises a more general question about the nature and function of the boundary between public and private spheres.[21] By inserting automatic enforcement functions into private spaces and activities, DRM self-help technologies elide the difference between public/rule-governed behavior and private behavior that is far more loosely circumscribed by applicable rules and social norms. Some offenses, most notably crimes against persons, are so severe that they may justify such elision. In other cases, however, looseness of fit between public rules and private behavior serves valuable purposes. Where privacy enables individuals to avoid the more onerous aspects of societal norms to which they may not fully subscribe, it promotes tolerance and pluralism. Where the precise contours of legal rules are unclear, or the proper application of legal rules to particular facts is contested, privacy shields a range of experimentation that furthers the value-balancing goals of public policy. DRM self-help technologies do not permit experimentation, and eliminate public policy and privacy alike from the calculus of infraction and enforcement. That these technologies, represent, at most, a novel form of “distributed totalitarianism” seems cold comfort.

III. Building Intellectual Privacy Into Law

Articulating legal principles for protecting the intellectual privacy interests implicated by DRM technologies is far more complicated than articulating the normative case for such protection. Normative theories are more supple than legal ones, which tend to move cautiously along well-trodden paths. Developing a legal theory of intellectual privacy for the information age requires an act of legal imagination. Because no single branch of legal doctrine supplies all of the elements necessary for effective protection of intellectual privacy, such a theory must synthesize elements from a variety of different legal traditions. It also must confront directly a problem that each of these doctrinal traditions has steadfastly avoided: determining what conditions should be necessary for an effective waiver of intellectual privacy if protection for intellectual privacy is to be meaningful. At both stages, the theory must be justified as an act of legal imagination. That is to say, it should be possible to show (capitulating at least partially to law’s inherent conservatism) that it at least does not differ too greatly from other such imaginative leaps.

A. Toward a Legal Privacy Standard for the Information Age

Many different strands of law bear to some degree on questions of intellectual privacy, but none is exactly developed to address the unique privacy problems created by DRM technologies. Several, however, have the potential to do so. The common law of privacy, with its emphasis on control over personal spaces and control over commercialization of image, can be reconfigured for the digital age by drawing on the policy and normative frameworks embodied in other privacy-regarding areas of law. In addition, because many information goods are also consumer goods, a more explicitly regulatory approach to privacy-invasive DRM technologies, grounded in principles of consumer protection law, can significantly improve levels of privacy protection for users of information goods.

1. Revitalizing Common Law Privacy Standards

The initial theory of common law privacy protection articulated by Warren and Brandeis was fairly flexible: a general “right to be let alone.”[22] The difficulty with this new right lay precisely in its generality and vagueness; without a more detailed specification, the right to be let alone could conceivably encompass almost any kind of unwanted attention. By the mid-twentieth century, and aided by legal scholarship seeking to subdue Warren and Brandeis’ unruly brainchild, the common law of privacy had congealed into four distinct torts.[23] The price of clarity, however, was stasis. Three of these torts are potentially applicable to the privacy problems created by DRM technologies, but all have remained firmly focused on the privacy problems of the predigital age. Yet each is potentially flexible enough to cover far more – if only courts become convinced that the expansion is warranted.

Current applications of the common law privacy torts do not encompass the sorts of incursions worked by DRM technologies. As noted in Part II, the tort of intrusion upon seclusion has targeted physical or audiovisual intrusions into private spaces.[24] No court has considered whether it similarly protects against the insertion of other kinds of sensors (e.g., DRM monitoring technologies), or sensors that report back to machines rather than to people, or technologies that drastically constrain behavior, but without reporting back. Each of these conclusions requires an additional step away from the traditional core of the tort. The fit between current conceptions of the common law privacy torts and informational privacy concerns is equally imperfect. The tort of unauthorized appropriation of name or likeness has focused primarily on misuse of proper names and pictorial images. So far, when asked to apply this tort to the digital “likenesses” generated by profiling and data mining activities, most courts have resisted.[25] The tort prohibiting embarrassing disclosure of private facts has generally been applied to redress disclosure of sexual or intimate information.[26] All three torts, however, are capable of a broader and more sensitive application.

Conceptual support for expansion of the common law privacy torts to cover electronic intrusion and monitoring can be found in policies derived from two bodies of law more finely attuned to intellectual privacy concerns: constitutional privacy law and copyright law. Compared with common law privacy rights, constitutional privacy rights manifest far greater concern with intellectual liberty. The liberty with which the drafters of the Constitution were concerned was liberty against government, and so constitutional protections for intellectual privacy have no direct application to the practices of private information providers. These protections are instructive nonetheless, for they reflect a set of values that our legal culture has identified as important and worth preserving. In particular, fourth and first amendment law supply principles designed to protect the spatial and informational attributes of intellectual privacy. Copyright law, meanwhile, implicitly presumes a degree of “breathing space,” and of anonymity, for users of intellectual goods. In different ways, then, each body of law intersects with and operationalizes (aspects of) the normative framework developed in Part II.

Application of the intrusion tort to DRM technologies finds parallels in a rapidly growing body of law that addresses the fourth amendment status of various types of remote information gathering. The federal courts have concluded that at least sometimes, such disembodied intrusions invade constitutionally protected privacy rights. Most recently, in Kyllo v. United States,[27] the Court held that extraction of heat signature information emanating from the defendant’s home constituted a search, and required a warrant. Kyllo does not address whether reporting back to a machine should count, yet on the Court’s reasoning there seems no reason why it should not. The search consists of the act of extraction, not what may or may not follow it. Rather, under Kyllo the dispositive factors were the fact that the extraction technology was “not in general public use” and the fact that it enabled access to “details of the home that would previously have been unknowable without physical intrusion.”[28]

As this brief discussion indicates, fourth amendment law has not embraced a spatialized understanding of intellectual privacy to the full extent suggested here. First, it remains unclear whether the strong privacy protection specified by the Kyllo Court is to be limited specifically to the home.[29] The majority’s brand of originalism supports this interpretation, but other approaches to constitutional interpretation and even other brands of originalism might not.[30] In delineating the legally cognizable scope of intellectual privacy interests, this is a particularly important question. Homes are but one kind of private space, and perhaps not even the most significant where intellectual activity is concerned.[31] Arguably, one’s desktop or laptop computer, personal data assistant, or portable media player sits at the center of the zone of intellectual privacy to which one is entitled, regardless of where in physical space it happens to be located.[32] Second, the “general public use” inquiry, like the “reasonable expectation of privacy” standard on which it builds, renders privacy a moving target. Eventually, the Court will need to confront the fact that the ultimate consequence of such a standard may be no privacy at all. Finally, by its own terms, the fourth amendment cannot even reach the question whether direct constraints, without any reporting back, invade a legally protectable privacy interest. Whether or not they do so, such constraints cannot constitute a “search.”

Here it is important to note – both for fourth amendment purposes and for insight into the lessons that the common law of privacy should draw from its constitutional cousin – that the text of the fourth amendment places intellectual privacy front and center. The amendment extends protection against warrantless search and seizure not simply to the home, but also to individuals’ “papers and effects.”[33] If individuals have no recourse for warrantless remote extraction of information from digital analogues to these items, this protection stands to lose much of its meaning. So too if widespread efforts to enshrine a new technology, such as DRM, as a commercial standard can displace privacy rights. Fourth amendment jurisprudence, like common law privacy jurisprudence, is just beginning to grapple with these and other difficult privacy problems presented by the digital age. Even so, its greater sensitivity to the intersections between spatial privacy and intellectual privacy is an important guidepost for courts in common law intrusion cases to follow, if they choose.

The argument that effective privacy protection should include control over the circumstances of intellectual consumption finds additional support within copyright law. The fair use doctrine, which sanctions certain acts of private copying, shields a range of actions that users might take in private spaces, including time- and space-shifting of copies, loading and reloading of digital files, and manipulation of digital content.[34] The first sale doctrine, which establishes the right to dispose of one’s copy of a work without any obligation to seek the copyright owner’s approval,[35] similarly rests on the belief that a copyright owner has no cognizable interest in a broad range of post-purchase user activities or in the spaces where they occur. More broadly, because copyright law does not give copyright owners the exclusive right to control all uses of their copyrighted works, it implicitly reserves to users the right to engage in conduct not encompassed by the statute.[36] Copyright does not, for example, encompass such acts as reading a copy of a book, viewing a copy of a movie, or listening to a copy of a musical recording that one owns; not coincidentally, these are all acts that ordinarily occur within private spaces.

The net effect of all of these rules is that copyright law traditionally has honored a version of the public-private distinction that is extremely robust.[37] Whether a provider of digital information is honoring or abusing this distinction should inform application of the common law intrusion tort, even to (at least some) DRM technologies that simply impose direct constraints on user behavior. From a copyright perspective the difference between reporting back and simple constraint is less relevant than the difference between public exploitation and private consumption. When deciding whether particular DRM constraints rise to the level of an actionable intrusion, courts should take this perspective into account.[38] The DMCA says nothing about its interaction with other federal or state privacy laws, just as it says nothing about its interaction with many other background rules of law, but that does not mean it negates them. The DMCA says nothing about its interaction with the background law of contract, either. That users are not authorized to circumvent a broader range of privacy-invasive measures does not mean that information providers have carte blanche to employ them. The most plausible explanation for §§ 1201(i) and 1205 is simply that interest groups brought specific problems to the drafting committees’ attention. The legislative history does not suggest that any of the relevant committees ever undertook a more thorough exploration of the privacy question.

Application of the appropriation and “embarrassing facts” torts to DRM monitoring technologies finds parallels in first amendment jurisprudence touching on intellectual privacy. First amendment cases involving the compelled disclosure of reading matter find intellectual activity quintessentially private because of the chilling effect on private expressive and political activity that might result from compelled disclosure of opinions and associations.[39] The chill may diminish when private compulsion replaces state compulsion, but it does not disappear. In the age of distributed databases, the pertinent fact is that a record of the activity exists, and may be acquired and used by either state or private parties.[40] On similar reasoning, the “embarrassing facts” tort should encompass disclosure of information about patterns of intellectual consumption. Arguably, the harms of such disclosure are at least as great as those resulting from disclosure of information about sexual activities and preferences, since it is the former rather than the latter upon which (a democratic/free) society relies to constitute its citizens. And if a profile of intellectual activities and preferences can chill expressive and associative conduct, it is hard to see why it should not be deemed a “likeness” – whether flattering or unflattering is beside the point – of the individual to whom it refers.

Further support for expansion of the appropriation tort to encompass transactional identity comes, paradoxically, from privacy’s commercial doppelganger, the common law right of publicity. Like the privacy tort of unauthorized appropriation, rights of publicity protect against unauthorized appropriation of names and likenesses. Rights of publicity typically are invoked to protect commercially valuable likenesses, while rights of privacy are not, but both theories seek to reserve control over commercial exploitation of identity to the individual with whom that identity is associated. Unlike courts hearing privacy cases, courts in publicity cases have generously construed the concept of “likeness,” extending protection to any attribute of personality that can reasonably be identified as belonging to the plaintiff.[41] Courts and commentators justify this expansion with reference to both the increasing value of (celebrity) identity and the many forms that identity can assume in the age of mass culture and advertising.[42] If it is true that manifestations of identity have become increasingly protean in the information age, there seems to be no good reason why the common law of privacy should not also recognize protectable attributes of identity in commercial profiles. Indeed, the case for such protection is far stronger than in the publicity context; data about one’s transactional history and preferences is far more directly bound up with identity than attributes that merely seek to trigger some mental association.

Finally, the same copyright rules that create a presumption of spatial privacy also provide strong implicit support for informational privacy claims directed toward DRM monitoring. Together, those rules effectively erect a strong presumption of anonymity around privileged uses. The functions and benefits of anonymity are clearest in the case of fair use. Fair use privileges a variety of activities that are deemed socially valuable, but to which private copyright holders might object. In other cases, the costs and delay involved in seeking permission may strike the would-be fair user as prohibitive, even if the overall social value resulting from the use would outweigh these costs.[43] Having to seek permission from the copyright holder ex ante would chill both types of uses; anonymity for fair users mitigates the twin problems of private censorship and high transaction costs, and allows society to receive the benefit of many controversial and/or spontaneous uses that otherwise would not occur.[44] But these benefits of anonymity accrue for other sorts of uses as well.

Synthesis of the intrusion, appropriation, and embarrassing facts torts with these insights, derived from conceptually related areas of law, would yield more expansive conceptions of actionable intrusion, appropriable identity, and sensitive personal information. This result is broadly consistent with the normative model of privacy developed in Part II, which focuses on control of access to self. It is also broadly consistent with the core policies underlying each tort: to preserve, respectively, individual control of space, identity, and “face.”

Why, though, should the common law of privacy make these leaps? For all the ingrained conservatism of the common law method, recognizing and responding to changing circumstances by redefining legally cognizable injury and responsibility is a central role of the courts. Many legal rules that we take for granted today simply did not exist forty or fifty years ago. One example is the law of strict products liability, under which an injured consumer may recover damages directly from the manufacturer of a defective product even if there is no privity of contract.[45] Another is the law of sexual harassment, which recognizes that sex-based hazing in the workplace can amount to discrimination in violation of federal law.[46] In each context, the courts gradually came to recognize that new forms of injury resulting from changed marketplace realities warranted new modes of redress.

In a similar fashion, courts can and should respond to new forms of injury enabled by the rise of digital network communications and the attendant transformations of commerce in information. In copyright circles, this point is hardly novel, but lawmakers and courts have focused their attention largely on new sources of injury to information providers. As these historical examples suggest, it is appropriate to focus, as well, on new sources of injury to information users, and doing so will not bring commerce in information screeching to a halt. The project of transforming existing doctrine to accommodate the unprecedented is itself firmly rooted in precedent.

There is one important caveats to this rather optimistic account of the potential for robust common law standards of intellectual privacy. Traditionally, common law privacy protections may be waived. As long as the contract is otherwise enforceable, one may consent to audio- or videotaping of the activities inside one’s home, or to commercial exploitation of one’s name or likeness, or to publication of sensitive information about one’s sexual habits. Because the privacy invasions effected by DRM technologies occur in the context of consensual commercial transactions, the mechanisms for establishing effective consent can easily be put in place.

Neither copyright law nor constitutional privacy law offers a clear way out here. Constitutional protections also can be waived. Copyright law, meanwhile, is silent about when parties may contract around the rights and limitations that it specifies. This silence has engendered an extensive scholarly debate about whether such contracts should be prohibited, under either a theory of preemption or one of misuse, as violating fundamental public policy.[47] Detailed consideration of those debates is outside the scope of this Article; for our purposes, the important point is that neither preemption nor misuse is well-suited to address the privacy problems stemming from DRM technologies. The fundamental public policy that both doctrines seek to preserve is the “copyright balance” between incentives and access. User privacy serves related purposes, and a decision striking down a particular contract provision might have the effect, as well, of promoting privacy, but privacy is not central to this inquiry. For a specifically privacy-regarding theory of contract’s limits, we must look elsewhere.

2. Consumer Protection Law and the Fair Information Practices

Although consumer protection law has not traditionally been viewed as a significant component of information policy in the United States, that is changing. In an era in which mass-distributed information goods are increasingly bundled with lengthy, complex licenses, the connections between consumer protection and information policy can no longer be ignored. Although the issue of privacy in intellectual consumption has not yet received specific attention, both the Federal Trade Commission (FTC) and intellectual property scholars have begun to focus more closely on these connections.[48] Where privacy is concerned, judge-made law and consumer protection regulation have complementary roles to play. While properly reformulated common law privacy torts can police the worst excesses of DRM, consumer protection law operating prospectively can set minimum standards of protection that all information providers must follow.

One advantage of a consumer protection approach to the terms of information access and use is that it allows policymakers to consider consumer welfare directly, rather than parsing out the implications of a statutory scheme (such as copyright) designed primarily to accomplish some other purpose. Whether this change in emphasis might translate into significant substantive protection for consumers depends on the prevailing standard for consumer well-being. United States consumer protection law is not particularly well tailored to safeguard the intellectual privacy of information users. Like the common privacy torts, however, it has the potential to be.

Consumer protection law in the United States has focused primarily, though not exclusively, on maximizing market-based indicia of consumer welfare. The FTC has jurisdiction to regulate “unfair or deceptive practices in or affecting commerce.”[49] In implementing this mandate, it has largely confined itself to policing deception, and has been much slower to implement measures designed to provide other sorts of protection to consumers who are adequately and accurately informed. Whatever the merits of this approach in other contexts, as an approach to privacy protection it is demonstrably inadequate. An extensive literature supports the conclusion that the idea of a well-functioning “market for privacy” is irremediably flawed.[50] In many transactions, retaining control of one’s personal information simply is not an option. Even when it is, pervasive and likely incurable information problems prevent individuals from evaluating the relevant tradeoffs.[51] More fundamentally, privacy tradeoffs involve incommensurable values, and the dignitary values at stake in decisions about privacy arguably are not an appropriate subject for market ordering at all.[52] For the reasons already discussed, this argument is particularly strong where intellectual privacy is concerned. Under the Clinton Administration, the FTC called without success for federal legislation establishing stronger protection for online privacy.[53] If the FTC wishes to play a role in safeguarding the intellectual privacy of information consumers, however, it can begin by rethinking its interpretation of its statutory mandate.

A somewhat more robust vision of informational privacy protection is embodied in guidelines issued in 1980 by the Organization for Economic Cooperation and Development, which outlined a set of Fair Information Practices (FIPs) based on eight principles: collection limitation, data quality, purpose specification, use limitation, transparency of information collection practices, security of stored data, individual participation, and accountability.[54] Although the U.S. played an important role in developing the FIPs, the FIPs have never been fully incorporated into U.S. law. In part, this is the result of sustained resistance by the information and direct marketing industries. In part, it is because the proceduralist understanding of consumer protection already enshrined within FTC practice pairs more comfortably with a version of fair information practices based simply on notice and consent. More faithful adherence to the FIPs would enhance the informational privacy of users of copyrighted works and other information goods.[55] The FTC has taken some steps in this direction, but only with respect to narrowly defined populations.[56] Extending protection to all consumers is appropriate in an age in which personal profiling increasingly extends far beyond purchases of durable goods to reach private intellectual activities.

Even with more rigorous and inclusive application of the FIPs, however, the problem of privacy in intellectual consumption is too complex to be resolved by data processing standards alone, for several reasons. First, the FIPs do not address spatial privacy, and so have nothing to say about the sorts of behavioral restrictions effected by DRM technologies. Thus, even scrupulous adherence to the FIPs would not address all of the privacy concerns discussed in Part II. Second, even with respect to information privacy, the FIPs do not establish minimum substantive thresholds for privacy protection. At most, they are designed to facilitate informed contracting and meaningful quality control by individuals who are the subjects of data transactions. Finally and relatedly, the FIPs do not address important threshold questions of contract validity. That is, they say nothing about whether waivers of privacy rights by consumers who are fully informed should nonetheless be subject to limits.

For consumer protection law to provide meaningful protection for intellectual privacy (or any other kind of privacy), the proceduralist standards embodied in the FIPs must be augmented by substantive privacy standards. Here the act of legal imagination consists in realizing that although the FTC has not traditionally involved itself in setting substantive standards of consumer protection, its mandate to address “unfair” trade practices is broad enough to encompass such a move. Put differently, a market-making conception of fairness is not the only possible definition of that term, nor the only sensible one. Where consumers cannot play on an equal footing with other market participants, it serves neither fairness nor markets to pretend they can.

In the context of informational privacy, one example of a substantive standard of fairness is the European Union’s data processing directive, which delineates certain kinds of information as sensitive and allows member states to place them off limits.[57] Similarly, if intellectual profiling is deemed to create unacceptable risk of harm to consumers, one might envision a regulation setting limits on the collection, use, retention, and trading of such information.[58] In the context of spatial privacy, an example of substantive privacy protection might be a regulation prohibiting certain kinds of electronic self-help,[59] or preserving a limited degree of freedom to space-shift digital files. By establishing these sorts of standards and enforcing penalties for violating them, consumer protection authorities can ensure that information consumers retain meaningful control over both the spatial and information dimensions of their own intellectual consumption. In addition, as I will discuss in Part IV, the law has an important role to play in ensuring that substantive protections for privacy are incorporated into the design of DRM technologies at the outset.

B. Contract and Intellectual Privacy as Fundamental Public Policy

The single greatest obstacle to effective legal protection of privacy in intellectual consumption is not imperfect fit with the available legal theories, but the fact that each available theory gives way to contract in many, if not all, circumstances. Many believe that this deference to contract is entirely appropriate. They observe that, from the information provider’s perspective, the greater power to withhold the transaction entirely logically includes the lesser power to impose conditions on the terms of access and use. From the individual user’s perspective, these conditions may diminish privacy, but users remain free to accept or reject the terms offered to them. Privacy advocates have persuasively argued that the argument from contract is far too simplistic, and ignores both marketplace realities and important non-market considerations. Thus far, however, the law has failed to translate these challenges into a workable legal theory capable of displacing contract when threats to privacy reach unacceptable levels.

Some challenges to contractual ordering of privacy rights focus on imperfections that are likely to prevent market mechanisms from working smoothly. These challenges fall into two general categories. First are procedural challenges to the validity of consent to online adhesion contracts waiving privacy. In the age of “clickwrap,” however, defects relating to consent are easily cured by requiring the consumer to pass through a screen displaying license terms and to indicate assent to those terms after having had the opportunity to review them.[60] A second set of objections to contractual ordering of privacy rights relates to issues of market power. If a dominant vendor has market power, it becomes harder to posit a meaningful level of competition to satisfy the full range of consumer preferences. Because the conventional form of this inquiry looks only to the power of individual market participants, however, and not to the market power that results from widespread adoption of standard form terms,[61] this argument has weight only in monopoly markets, and (as a result) very little weight in most markets for online information goods.

Both types of argument from market imperfection, however, fit comfortably within a larger conceptual framework that presumes the rightness of market ordering if only some defect or other could be brought under control. Neither challenges the baseline presumption in favor of contractual ordering in properly functioning markets. As a result, each rapidly becomes mired in the details of this or that clickwrap procedure, or this or that market practice. The more fundamental question – whether market ordering of privacy rights makes sense at all – remains obscured. It is not terribly surprising, then, that these sorts of arguments have failed to generate the impetus for meaningful reform of the legal rules governing waiver of privacy rights.

Other challenges to contract step outside the market framework, and argue that even in perfectly functioning markets, contract would be ineffective to preserve privacy, or to do so fairly.[62] As discussed in Part III.A.2, above, some of these arguments rest on the premise that in the modern mass marketplace, consumer choice about privacy is fictitious; others point to the insoluble information problems that consumers confront in assessing privacy tradeoffs; and still others reject a priori the notion that market resolution of privacy policy is appropriate.[63] On any of these views, the problem is not market failure, but rather a more systemic failure of markets.

It is a measure of the degree to which both academic and policy debates have been captured by the rhetorics of markets and private ordering that these arguments receive comparatively little attention. In the current climate, arguments from human dignity seem both insufficiently rigorous and vaguely passe. Yet the reluctance to address privacy in non-market terms is puzzling, for two reasons. As Jessica Litman has pointed out (and as privacy advocates “in the trenches” have always known), that is the way that ordinary people think about privacy.[64] Ordinary people – not academics, technologists, science fiction writers, or other members of the cyber-literati – react to abuses of privacy with outrage and a sense of betrayal, and feel that commercial dealings should be accompanied by privacy obligations.[65] That this outrage often does not translate into meaningful market resistance should not surprise us[66]; if markets for privacy are inherently dysfunctional, there is no reason to expect this result.

If one asks, instead, the more general question whether there are circumstances in which public policy overrides contract, one discovers that many public policy-based limits on contract exist. In fact, the proposition becomes much less remarkable than the rhetoric of current privacy debates makes it seem. Most people agree that there are some public policies that should not be altered by contract. Perhaps the best example is the general policy that one may not contract into a state of slavery, but there are many other, less dramatic examples. One is the rule that one may not sell one’s organs for transplant, research, or any other use.[67] Two additional examples are the rules that one may not contract out of medical malpractice liability or strict products liability by asserting one’s willingness to risk injury in return for a lower price.[68] Still another, more recent example is set forth in a New York trial court’s ruling enjoining a software developer from forbidding licensees to publish critical reviews of its products.[69] In each of these situations, the question whether the “free market” might equilibrate in a way that preserves the default rule is considered irrelevant.

This brief list illustrates two salient points about the sorts of public policies that are considered “important” enough to trump contract. First, these policies bolster noneconomic values that run the gamut from bodily integrity to freedom of expression to human dignity and self-determination. Privacy in general, and intellectual privacy in particular, serve values that fall comfortably within this spectrum. Second and equally important, the appeal to public policy is not an appeal to logic or political theory, but to visceral notions of fairness. For privacy concerns to trump contract, privacy advocates must establish not only that privacy values are similar in kind to other public values that society has sought to preserve, but also that they are similarly compelling. Once convinced of this, courts could quite easily develop rules limiting privacy waivers just as they have done in other contexts.

At bottom, the argument for limiting waiver of intellectual privacy rights is straightforward, and builds upon the argument in Parts II and III.A, above, about why intellectual privacy is important and why the law should recognize harms to intellectual privacy in the first instance. Sophistry about markets and market failures aside, intangible invasions of intellectual privacy are capable of causing great harm to individuals, and of substantially undermining shared, nonmonetizable values. Such invasions compromise rights of self-determination and undermine human dignity by eliminating the “breathing space” for intellectual development. A decision to promote these values in the law of “privacy” while simultaneously enabling easy evasion of accountability via “contract” would be nothing short of perverse. Taking these intangible harms seriously requires more.

IV. Building Intellectual Privacy Into Code

Although legal sanctions for invasion of intellectual privacy are essential to guarantee respect for the intellectual privacy rights of information users, as a means of ensuring effective protection for all users, both judicial and regulatory sanctions are second-best strategies. A far more effective method of ensuring that information users actually enjoy the privacy to which they are entitled would entail building privacy into the design of DRM technologies in the first instance. In such a world, legal protection for intellectual privacy would serve as backdrop to more privacy-regarding conduct by (most) providers of information goods, and would simply serve to police deviations from an accepted range of privacy protection. Taking privacy into account at the outset requires a different approach to designing DRM technologies, and also requires a process for ensuring that more privacy-protective DRM technologies are put in place.

A. Value-Centered Design for DRM

The notion of value-centered design is an outgrowth of the interdisciplinary study of science, technology, and society. Careful attention to the social embeddedness of technologies reminds us that technologies themselves are social artifacts; they constitute and are constituted by social values and interests.[70] In the context of DRM, this insight suggests that design for maximum control is but one direction that a DRM infrastructure could take. Alternatively, one might imagine developing a design process devoted to identifying the full range of values, both private and public, implicated in DRM design, and to operationalizing DRM in a way that preserves important public values. Such a value-centered design process for DRM technologies would seek, among other things, to create rights management infrastructures for information goods that respect and seek to preserve user privacy.[71] Such an infrastructure would have three components, which map to the three types of privacy invasion discussed in Part II.

The first component of value-centered design for DRM involves development of flexible restrictions that minimize direct constraints on intellectual consumption within private spaces. The idea that functionality restrictions might be designed to preserve flexibility for private access and copying is hardly novel. One example is the serial copy management system mandated by the Audio Home Recording Act, which allows the production of perfect first-generation copies but causes significant quality degradation in subsequent generations.[72] Another example is the Digital Millennium Copyright Act’s requirement that analog videocassette recorders be designed to allow consumers to time-shift some kinds of television programming.[73] Elsewhere, Dan Burk and I have argued that restrictions similar to these are necessary to preserve basic user privileges established under copyright law, such as fair use.[74] Flexible restrictions on the functionality of digital copies also would operate to preserve user privacy, for similar reasons. Such restrictions rest on the presumption that an information provider has no legitimate interest in controlling or even knowing about certain types of uses within private spaces.

Value-centered design for DRM also would build in limits on monitoring and profiling of individual users. Because most businesses need to collect and retain some information about their customers to manage orders, payments, and deliveries, technological limits on data collection and use cannot fully substitute for other, human-implemented rules. Nonetheless, DRM systems may be designed either to minimize or to maximize data collection, retention, extraction and use. To preserve the intellectual privacy of information users, DRM design should incorporate a minimization principle.[75] In the cases where real-time monitoring of user conduct is deemed to provide some significant non-privacy-related benefit,[76] designers should consider whether the desired benefit can be achieved capturing the precise identity of the user, or without tying users to content. If not, and if the implementation ultimately chosen must reflect a choice between the benefit and user privacy, that choice should be made explicitly and documented, so that later designers, regulators, and courts can understand the tradeoffs involved.

Finally, a privacy-regarding DRM infrastructure would implement limitations on self-help. This might mean, for example, that digital files could not be programmed to self-destruct, or to deny access entirely, upon detecting an impermissible action.[77]

These proposals are necessarily quite general. Whether they would operate to guarantee meaningful levels of privacy for information users would depend upon the specific details of their implementation. Nor are these suggestions necessarily the only or the best ones; an expert in the relevant technological fields could undoubtedly think of others. The point is simply that a focus on value-centered design exposes “DRM” as a concept that is susceptible of a wide range of meanings. Understanding the DRM design process as (necessarily) value-driven is an essential first step to ensuring that in the future, design priorities shift to include a broader range of values.

B. Creating a Value-Centered Design Process

Identifying the possibility of value-centered design for privacy is only half the battle. For privacy-regarding DRM technologies to move from the pages of academic articles onto of the drawing board and ultimately into the marketplace, those who participate in or underwrite real-world design processes need incentives to expand their frames of reference. Law has a role to play here as well, although it is a very different role from that discussed in Part III. Law’s role in structuring DRM standard-setting processes is to ensure that the formulation of technical standards by market actors takes public values, including privacy values, into account.

If, as several advocacy organizations have urged, the law were to specify a “bill of rights” for users of information goods, this would constrain DRM development initiatives to focus on public values as well as the private ones.[78] In particular, rights of intellectual privacy could be specified at a sufficiently high level of generality to avoid dictating the choice of technical standards, while still conveying the necessary information about the substance of the protection to be afforded. Thus, following the model set forth above, rights of intellectual privacy would include: the right not to be subjected to intrusive constraints on the use of intellectual goods within private spaces; rights against monitoring of intellectual consumption and profiling based on intellectual preferences; and the right not to be subjected to electronic self-help that would disable access to lawfully acquired information goods. Development of technical standards and processes to effectuate these rights would be the content industries’ affair.

Vigilant defenders of market ordering will object that this proposal improperly injects government into a process – standards development – that is quintessentially of, by, and for the market. It takes but a moment’s reflection to see that this objection is simply the first cousin once removed of the old argument for market ordering of privacy rights. If the first-order “market for privacy” cannot accurately reflect the variety of values placed on privacy, it is difficult to imagine how a second-order market for privacy standards, derived by inference from the first-order market for privacy, could possibly do so. Even assuming that the first-order market for privacy actually worked, and that the average consumer could easily master the complexity of DRM standards terminology, market processes are not well suited to enable consumers to exert positive, as opposed to negative, influence on the design of technical standards. Consumers can refuse to buy, or can switch from one provider to another, but there are no mechanisms to allow consumers to communicate as a prospective matter the precise level of functionality that they want. And because DRM technologies are network technologies,[79] it will become increasingly difficult for dissenting consumers to opt out.

DRM standards processes offer an opportunity for more reflective participation in the debate over DRM, but still are not good vehicles for the incorporation of public values into DRM design. To the average consumer of information goods, standards processes are arcane and relatively inaccessible proceedings. Organizations representing consumers and other noncommercial interests have begun to take an interest in DRM standard-setting.[80] At present, however, their participation in these processes is largely on the sufferance of the content industries. Not all standards processes include consumer representation, and even in those that do, there is no assurance that consumer grievances, once aired, will actually make their way into the standards that are brought to market.[81]

For a value-centered design process to succeed, some actor external to the market for DRM technologies must identify and maintain the centrality of the relevant public values. I do not intend to suggest that the law should mandate the content of technical standards for DRM technologies, or that government actors would be good at supervising such a process. Government can be rather good, though, at mandating non-technical standards.[82] In the predigital world, we might call these non-technical standards simply “rights.” In the digital age, rights state (among other things) the values that technical standards should be designed to enable. [EXPAND A BIT HERE]

V. Conclusion

TO COME. [NOTES: The key to preserving the benefits of intellectual privacy is that both right-holders and the law must surrender some control over user behavior. But that’s exactly what freedom is. These questions need to be asked while design and implementation are still on the drawing board; it will be much easier to design for privacy now than to retrofit privacy-invasive designs later.]

---

*Professor of Law, Georgetown University Law Center. Internet: I thank Andrew Crouse for his able research assistance.

[1]See, e.g., Judith Jarvis Thomson, The Right to Privacy, in Ferdinand David Schoeman, ed., Philosophical Dimensions of Privacy: An Anthology 272 (1984).

[2]See Ferdinand David Schoeman, Privacy: Philosophical Dimensions of the Literature, in Ferdinand David Schoeman, ed., Philosophical Dimensions of Privacy: An Anthology 1 (1984).

[3]See, e.g., Katz v. United States, 389 S. Ct. 347 (1967) (holding that a person has a reasonable expectation of privacy while using a public telephone booth); Doe by Doe v. B.P.S. Guard Services, Inc., 945 P.2d 1422 (8^th Cir. 1991) (holding that surreptitious videotaping of fashion models in their dressing room was an invasion of privacy); Benitez v. KFC Nat’l Mgmt. Co., 714 N.E.2d 1002 (Ill. App. 1999) (holding that female employees’ allegations that employer spied on them through hole in ceiling of women’s bathroom stated claim for invasion of privacy); Harkey v. Abate, 346 N.W.2d 74 (Mich. App. 1983) (holding that installation of hidden viewing device in public bathroom at skating rink invaded privacy); People for Ethical Treatment of Animals v. Bobby Berosini, Ltd., 895 P.2d 1269 (Nev. 1995) (add parenthetical). But see Elmore v. Atlantic Zayre, Inc., 341 S.E.2d 905 (Ga. App. 1986) (holding that rights of privacy in store bathrooms may be outweighed by store’s interest in deterring crime); Hougum v. Valley Memorial Homes, 574 N.W.2d 812 (N.D. 1998) (no invasion of privacy where employee only unintentionally observed man masturbating in public bathroom).

[4]Cf. Erving Goffman, The Presentation of Self in Everyday Life (1959); Alan F. Westin, Privacy and Freedom (1970).

[5]See, e.g., cites.

[6]See, e.g., Anita L. Allen, Coercing Privacy, 40 Wm. & Mary L. Rev. 723, 754-55 (1999); Julie E. Cohen, Examined Lives: Informational Privacy and the Subject as Object, 52 Stan. L. Rev. 1373, 1424-28 (2000); Julie E. Cohen, A Right to Read Anonymously: A Closer Look at “Copyright Management” in Cyberspace, 28 Conn. L. Rev. 981, 1006-14 (1996); Ruth Gavison, Privacy and the Limits of Law, 89 Yale L.J. 421 (1980); Seth F. Kreimer, Sunlight, Secrets, and Scarlet Letters: The Tension Between Privacy and Disclosure in Constitutional Law, 140 U. Pa. L. Rev. 1, 59-71 (1991).

[7]Cf. Jeffrey Rosen, The Unwanted Gaze: The Destruction of Privacy in America (2000) (arguing that privacy protects the individual interest in not being judged “out of context”); Radhika Rao, A Veil of Genetic Ignorance? Protecting Privacy as a Mechanism to Ensure Equality, __ Hastings L.J. (forthcoming 2003) (arguing that privacy is grounded in equality interests).

[8]See Video Privacy Protection Act of 1988, 18 U.S.C. § 2710; [title of cable act], 47 U.S.C. §551; Julie E. Cohen, A Right to Read Anonymously: A Closer Look at “Copyright Management” in Cyberspace, 28 Conn. L. Rev. 981, 1031 n.213 (1996) (collecting state statutes safeguarding the privacy of library patrons).

[9]See P.J. Huffstutter & Jon Healey, Suit Filed Against Record Firms, L.A. Times, June 14, 2002; Brenda Sandburg, Milberg Weiss Files Suit Over CDs With No-Copy Technology, The Recorder, June 17, 2002; Amy Harmon, CD-Protection Complaint Is Settled, N.Y. Times, Feb. 25, 2002, at C8.

[10]See John Borland, Studios Race to Choke DVD Copying, CNET News.com, Feb. 4, 2002; Matt Lake, How It Works: Tweaking Technology to Stay Ahead of the Film Pirates, N.Y. Times, Aug. 2, 2001; Doug Mellgren, Acquittal in DVD Decoding: Norwegian Teen Created Program So He Could View Film on Computer, Charlotte Observer, Jan. 8, 2003.

[11]See, e.g., Schoeman, supra note __; add others.

[12]Cf. Lawrence Lessig, Code and Other Laws of Cyberspace ___ (1999); Joel R. Reidenberg, Lex Informatica: The Formulation of Information Policy Rules Through Technology, 76 Tex. L. Rev. 553 (1998).

[13]For some privacy skeptics, the failure of privacy scholars to agree on a single definition of privacy is a fundamental weakness. Other scholars, however, contend that recourse to multiple, sometimes overlapping definitions of privacy is entirely reasonable, and does not weaken the case that “privacy” interests exist. See Daniel J. Solove, Conceptualizing Privacy, 90 Calif. L. Rev. 1087 (2002) (suggesting a pragmatic, family-of-concepts approach to privacy).

[14]See, e.g., cites – MacKenzie; others.

[15]See Specht v. Netscape Communications Corp., 306 F.3d 17 (2d Cir. 2002) (affirming denial of defense motion to compel arbitration in case alleging invasion of privacy by use of browser “plug-in” to monitor online activity); In re RealNetworks, Inc. Privacy Litigation, 2000 WL 631341 (N.D. Ill. 2000) (rejecting arguments by intervenor in proceeding to compel arbitration of privacy claims regarding media player software that monitored and stored information about users’ electronic communications); cf. In re Pharmatrak, Inc. Privacy Litigation, 220 F. Supp. 2d 4 (D. Mass. 2002) (holding that Pharmatrak was not liable under federal electronic communications, wiretap, or computer fraud laws for using “cookies” to collect personal information about web site users); In re DoubleClick, Inc. Privacy Litigation, 154 F. Supp. 2d 497 (S.D.N.Y. 2001) (same).

[16]See Mark Prigg & Avril Williams, Spies Behind Your Screen, The Times (London), Aug. 6, 2000; add others.

[17]For good discussions of profiling and its uses, see Oscar H. Gandy, Jr., The Panoptic Sort: A Political Economy of Personal Information (1993); Jeff Sovern, Opting In, Opting Out, or No Options at All: The Fight for Control of Personal Information, 74 Wash. L. Rev. 1033 (1999).

[18]See Unif. Cptr. Info. Transactions Act (UCITA) §§ 605, 815-816 (as amended 2002); UCITA §§ 605, 815-816 (as amended 2001); UCITA §§ 605, 815-816 (1999); Unif. Commercial Code (UCC) §§ 2B-310 and 2B-715, Reporter’s Note 3 (Aug. 1, 1998 Draft); UCC §2B-310, -716 (Apr. 1, 1998 Draft); UCC §2B-310, -716 (Feb. 1998 Draft); see also Julie E. Cohen, Copyright and the Jurisprudence of Self-Help, 13 Berkeley Tech. L.J. 1089 (1998).

[19]See Ruth Gavison, Privacy and the Limits of Law, 89 Yale L.J. 421 (1980).

[20]For this reason, it may make more sense to view these activities through the lens of objectification rather than that of accessibility. See Julie E. Cohen, Examined Lives: Informational Privacy and the Subject as Object, 52 Stan. L. Rev. 1373 (2000).

[21]“Public” and “private” are terms with multiple meanings. I use “private” here not to denote non-state activities, but simply to denote spaces not open to the general public and behaviors not intended for the general public, including private intellectual activities, and “public” to include conduct that occurs outside these realms.

[22]Samuel D. Warren & Louis D. Brandeis, The Right to Privacy, 4 Harv. L. Rev. 193 (1890).

[23]See William L. Prosser, Privacy, 48 Calif. L. Rev. 338 (1960); W. Page Keeton, et al., Prosser and Keeton on Torts §117 (5^th ed. 1984); Restatement (2d) of Torts §652A (1977).

[24]See, e.g., Miller v. Brooks, 472 S.E.2d 350 (N.C. App. 1996) (holding that placing a video camera in plaintiff’s bedroom and going through his mail could constitute intrusion upon seclusion); Association Services, Inc. v. Smith, 549 S.E.2d 454 (Ga. App. 2001) (holding that trespassing upon private property while conducting surveillance could constitute intrusion upon seclusion); Clayton v. Richards, 47 S.W.3d 149 (Tex. App. 2001) (holding that placing a video camera in plaintiff’s bedroom could constitute intrusion upon seclusion).

[25]See, e.g., Dwyer v. American Express Co., 652 N.E.2d 1351 (Ill. App. Ct. 1995) (holding that credit card company did not appropriate cardholders’ names or likenesses by renting lists of their names characterized by purchasing patterns); Avrahami v. U.S. News & World Report, Inc., 1996 WL 1065557 (Va. Cir. Ct. 1996) (holding that media company did not appropriate customer’s name or likeness by selling information about him). There are some signs, however, that this resistance may be lessening. See Weld v. CVS Pharmacy, Inc., 1999 WL 494114 (Mass Super. Ct. 1999) (denying defense motion for summary judgment on claim that it appropriated plaintiffs’ names and likenesses by selling information about them); Bodah v. Lakeville Motor Express, Inc., 649 N.W.2d 859 (Minn. Ct. App. 2002) (holding that employees stated a claim for appropriation of name or likeness where employer transmitted their social security numbers to third parties).

[26]See, e.g., Doe v. Mills, 536 N.W.2d 824 (Mich. App. 1995) (holding that plaintiff stated prima facie case of public disclosure of embarrassing facts where anti-abortion protesters displayed her name outside an abortion clinic); Y.G. v. Jewish Hosp. of St. Louis, 795 S.W.2d 488 (Mo. App. 1990) (holding that plaintiffs stated claim for public disclosure of embarrassing facts where information about their participation in hospital in vitro fertilization program was televised).

[27]533 U.S. 27 (2001); see also United States v. Karo, 468 U.S. 705 (1984). But see Smith v. Maryland, 442 U.S. 735 (1979). (add parentheticals).

[28]Id. at 40.

[29]See Andrew Riggs Dunlap, Fixing the Fourth Amendment with Trade Secret Law: A Response to Kyllo v. United States, 90 Geo. L.J. 2175, 2190 (2002).

[30]Cf. Lawrence Lessig, Fidelity in Translation, 71 Tex. L. Rev. 1165 (1993); Lessig, supra note __; Michael Adler, Note, Cyberspace, General Searches, and Digital Contraband: The Fourth Amendment and the Net-Wide Search, 105 Yale L.J. 1093 (1996); Dunlap, supra note __.

[31]See Dunlap, supra note __, at 2187 (“Modern America is defined by the mobility of its people and their information.”).

[32]Perhaps for this reason, government agents appear to believe that a warrant is required for searches of these items. See United States v. Runyan, 290 F.3d 223 (5^th Cir. 2002); Muick v. Glenayre Elec., 280 F.3d 741 (7^th Cir. 2002); United States v. Triumph Capital Group, Inc., 2002 WL 314877 (D. Conn. 2002). But cf. Aronson v. Spectrum, 767 A.2d 564 (Pa. Super. Ct. 2001) (holding that telecommunications company did not intrude upon customers’ seclusion by allowing third parties to access their account information).

[33]U.S. Const. amend. IV; see Dunlap, supra note __, at 2190-93.

[34]17 U.S.C. § 107; see Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417 (1984); Mattel, Inc. v. Pitt, ___ F. Supp. 2d ___ (S.D.N.Y. 2002); cf. Recording Indus. Ass’n of Am. v. Diamond Multimedia Sys., 180 F.3d 1072, 1079 (9^th Cir. 1999) (add parenthetical).

[35]17 U.S.C. §109(a).

[36]In this respect, the fair use doctrine is poorly named; the term “fair use” tends to suggest that if some uses of copyrighted works are fair, then all other uses must be unfair. Fair use and other copyright limitations are not outer limits on permissible uses of copyrighted works and/or the things embodying them. They are simply outer limits on a copyright owner’s statutory rights. Uses not covered by any of those rights, such as reading a copy of a book that one owns, are reserved to users whether or not the fair use doctrine would apply to them.

[37]See Paul Goldstein, Copyright’s Highway: From Gutenberg to the Celestial Jukebox (1994) (but argues that copyright should extend its reach into private spaces).

[38]It may be argued that the Digital Millennium Copyright Act’s (DMCA) protections for DRM technologies threaten to change rather substantially, and as a matter of federal law, the degree of privacy to which users of intellectual goods are entitled. In fact, the language of the DMCA supports the opposite conclusion: Congress did not intend the DMCA to negate the intellectual privacy of information consumers.

An exception to the DMCA’s anti-circumvention provision authorizes users of copyrighted works to circumvent technical measures capable of collecting or disseminating information about their “online activities.” See 17 U.S.C. § 1201(i). Under this provision, users appear free to subvert certain types of DRM monitoring. In addition, a special savings provision of the statute expressly preserves federal and state laws protecting individual privacy “in connection with the individual’s use of the Internet.” 17 U.S.C. §1205 (“Nothing in this chapter abrogates, diminishes, or weakens the provisions of, nor provides any defense or element of mitigation in a criminal prosecution or civil action under, any Federal or State law that prevents the violation of the privacy of an individual in connection with the individual’s use of the Internet.”). This provision is probably best interpreted as preserving information providers’ obligations under the federal Electronic Communications Privacy Act and analogous state laws; thus, for example, a software company caught monitoring customers’ use of its e-mail program could not claim that the DMCA allows it to do so.

[39]See Denver Area Educational Telecommunications Consortium, Inc. v. FCC, 116 S. Ct. 2374, 2390-96 (1996) [needU.S. cite]; Stanley v. Georgia, 394 U.S. 557, 563-66 (1969); Schneider v. Smith, 390 U.S. 17, 24-25 (1968); Lamont v. Postmaster General, 381 U.S. 301, 307 (1965); Fabulous Associates, Inc. v. Pennsylvania Public Utility Comm., 896 F.2d 780, 785 (3d Cir. 1990); see also; Gibson v. Florida Legislative Investigation Comm., 372 U.S. 539 (1963); Bates v. City of Little Rock, 361 U.S. 516 (1960); NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 460-62 (1958). See generally Julie E. Cohen, A Right to Read Anonymously: A Closer Look at “Copyright Management” in Cyberspace, 28 Conn. L. Rev. 981 (1996).

[40]See Tattered Cover, Inc. v. City of Thornton, 44 P.3d 1044 (Colo. 2002); In re Grand Jury Subpoena to Kramerbooks & Afterwords, Inc., 26 Med. L. Rptr. 1599 (D.D.C. 1998); see also Julie E. Cohen, Privacy, Technology, and Ideology: A Response to Jeffrey Rosen, 89 Geo. L.J. 2029, 2033 (2001).

[41]See, e.g., White v. Samsung Elec. Am., 971 F.2d 1395 (9^th Cir. 1992) (game show hostess’s gown and game show setting), petition for reh’g and reh’g en banc denied, 989 F.2d 1512 (9^th Cir.), cert. denied, 508 U.S. 951 (1993); Waits v. Frito-Lay, Inc., 978 F.2d 1093 (9^th Cir. 1992) (imitation of singer’s distinctive voice and singing style); Midler v. Ford Motor Co., 849 F.2d 460 (9^th Cir. 1988) (same); Carson v. Here’s Johnny Portable Toilets, Inc., 698 F.2d 831 (6^th Cir. 1983) (talk show host’s “trademark” slogan); Motschenbacher v. R.J. Reynolds Tobacco Co., 498 F.2d 821 (9^th Cir. 1974) (race car driver’s distinctively decorated car).

[42]See Roberta Rosenthal Kwall, Fame, 73 Ind. L.J. 1 (1997); add others.

[43]See Julie E. Cohen, Lochner in Cyberspace: The New Economic Orthodoxy of “Rights Management”, 97 Mich. L. Rev. 462 (1998); Lydia Pallas Loren, Redefining the Market Failure Approach to Fair Use in an Era of Copyright Permission Systems, 5 J. Intell. Prop. L. 1 (1997); cf. Mark A. Lemley, The Economics of Improvement in Intellectual Property Law, 75 Tex. L. Rev. 989 (1997).

[44]See Dan L. Burk & Julie E. Cohen, Fair Use Infrastructure for Rights Management Systems, 15 Harv. L.J. & Tech. 41 (2001).

[45]See Escola v. Coca-Cola Bottling Co. of Fresno, 150 P.2d 436, 461 (Cal. 1944) (Traynor, J., concurring); Sheward v. Virtue, 126 P.2d 345 (Cal. 1942); State Farm Mut. Auto. Ins. Co. v. Anderson-Weber, Inc., 110 N.W.2d 449 (Iowa 1961); Carter v. Yardley & Co., 64 N.E.2d 693 (Mass. 1946); McCormack v. Hankscraft Co., 154 N.W.2d 488 (Minn. 1967); MacPherson v. Buick Motor Co., 111 N.E. 1050 (N.Y. 19__); Ritter v. Narragansett Elec. Co., 283 A.2d 255 (R.I. 1971).

[46]See Meritor Savings Back v. Vinson, 477 U.S. 57 (1986); Bundy v. Jackson, 641 F.2d 934 (D.C. Cir. 1981); Tomkins v. Public Service Elec. & Gas Co., 568 F.2d 1044 ( 3d Cir. 1977); Berkman v. City of New York, 580 F. Supp. 226 (S.D.N.Y. 1983).

[47]See, e.g., Mark A. Lemley, Beyond Preemption: The Law and Policy of Intellectual Property Licensing, 87 Cal. L. Rev. 111 (1999); David Nimmer et al., The Metamorphosis of Contract Into Expand, 87 Calif. L. Rev. 17 (1999); J.H. Reichman & Jonathan A. Franklin, Privately Legislated Intellectual Property Rights: Reconciling Freedom of Contract with Public Good Uses of Information, 147 U. Pa. L. Rev. 875 (1999); David A. Rice, Public Goods, Private Contract, and Public Policy: Federal Preemption of Software License Prohibitions Against Reverse Engineering, 53 U. Pitt. L. Rev. 543 (1992).

[48]See U.S. Federal Trade Comm’n, Competition and Intellectual Property Law and Policy in the Knowledge-Based Economy (Feb.-Nov. 2002), <>; U.S. Federal Trade Comm’n, Warranty Protection for High-Tech Products and Services (Oct. 26-27, 2000), <>.

[49]15 U.S.C. § 45(a)(1)).

[50]See, e.g., Gandy, supra note __; Cohen, supra note __; A. Michael Froomkin, Flood Control on the Information Ocean: Living with Anonymity, Digital Cash, and Distributed Databases, 15 J.L. & Com. 395, 492 (1996); Paul M. Schwartz, Privacy and Democracy in Cyberspace, 52 Vand. L. Rev. 1609 (1999); Paul M. Schwartz, Privacy and the Economics of Personal Health Care Information, 76 Tex. L. Rev. 1, 47-51 (1997); Sovern, supra note __.

[51]See Cohen, supra note __, at 1397-99; Froomkin, supra note __, at 492; Schwartz, Personal Health Care Information, supra note __, at 47-51; Sovern, supra note __.

[52]See Cohen, supra note __; Schwartz, Privacy and Democracy, supra note __. For this reason, it may make sense to conclude that the law should protect (some aspects of) privacy even for individuals who would cheerfully trade it away. See Anita L. Allen, Coercing Privacy, 40 Wm. & Mary L. Rev. 723 (1999); Cohen, supra note __.

[53]See U.S. Federal Trade Commission, Privacy Online: Fair Information Practices in the Electronic Marketplace (2000).

[54]See O.E.C.D., Recommendation of the Council Concerning Guidelines Governing the Protection of Privacy and Transborder Flows of Personal Data, O.E.C.D. Doc. C(80) 58 final (1980), <http://www.oecd.org/EN/document/0,,EN‑document‑43‑1‑no‑24‑10255‑43,00.html>.

[55]For that matter, it would also enhance the functioning of markets in personal information by ensuring that personal information is accurate and that data processing operations more completely internalize their costs.

[56]See U.S. Federal Trade Comm’n, Children’s Online Privacy Protection Rule, 64 Fed. Reg. 59,887 (Nov. 3, 1999), codified at 16 C.F.R. § 312; see also U.S. Dep’t of Commerce, Safe Harbor Overview (establishing guidelines for U.S. companies that process personally identifying information relating to European Union citizens, and vesting enforcement authority with the FTC for most industries), <>.

[57]Council Directive 95/46/EC on the protection of individuals with regard to the processing of personal data and on the free movement of such data, 1995 O.J. (L. 281) 31, art. 8.

[58]Such a regulation might be modeled on the Video Privacy Protection Act of 1988, 18 U.S.C. § 2710, or on state library privacy statutes. See supra note __.

[59]Such a regulation would also have the beneficial effect of resolving the ongoing debate among the drafters of UCITA. See supra note __.

[60]See, e.g., Unif. Cptr. & Info. Transactions Act §___; ProCD, Inc. v. Zeidenberg, 86 F.3d 1447 (7^th Cir. 1996). This is not to make light of what commentators rightly identify as a paradigm shift in prevailing understandings of the sort of consent required to create a binding contract. See, e.g., Margaret Jane Radin, Humans, Computers, and Binding Commitment, 75 Ind. L.J. 1125 (2000). But that paradigm shift resulted from the rise of consumer mass markets decades ago. Technologies for indicating “consent” online simply underscore what we already know to be true: that in mass markets, the idea of a “meeting of minds” is little more than a pleasant fiction.

[61]See Victor P. Goldberg, Institutional Change and the Quasi-Invisible Hand, 17 J.L. & Econ. 461, 468 n.15, 484-91 (1974); Friedrich Kessler, Contracts of Adhesion – Some Thoughts About Freedom of Contract, 43 Colum. L. Rev. 629 (1943); Todd D. Rakoff, Contracts of Adhesion: An Essay in Reconstruction, 96 Harv. L. Rev. 1173 (1984); W. David Slawson, Standard Form Contracts and Democratic Control of Law-Making Power, 84 Harv. L. Rev. 529, 538-42 (1971); William T. Vukowich, Lawyers and the Standard Form Contract System: A Model Rule That Should Have Been, 6 Geo. J. Legal Ethics 799, 800-11 (1993); see also Robert P. Merges, Intellectual Property and the Costs of Commercial Exchange: A Review Essay, 93 Mich. L. Rev. 1570, 1611-13 (1995) (examining standard form terms within the narrower context of antitrust-style market power).

[62]See Cohen, supra note __; Paul Schwartz.

[63]See supra pp. ___.

[64]Jessica Litman, Information Privacy/Information Property, 52 Stan. L. Rev. 1283 (2000).

[65]See id. at ___; Laura J. Gurak, Persuasion and Privacy in Cyberspace: The Online Protests Over Lotus Marketplace and the Clipper Chip (1997).

[66]The lack of market resistance by consumers is routinely invoked by privacy opponents as purportedly demonstrating a lack of genuine public concern with privacy. See, e.g., cites.

[67]See, e.g., 42 U.S.C. § 274e; Newman v. Sathyovaglswaran, 287 F.3d 786 (9^th Cir. 2002). [add more]

[68]See Tunkl v. Regents of Univ. of Cal., 383 P.2d 441 (Cal. 1963) (holding that required agreement releasing hospital from malpractice liability was void as against public policy); Ash v. New York Univ. Dental Ctr., 564 N.Y.S.2d 308 (App. Div. 1990) (same); Westlye v. Look Sports, Inc., 22 Cal. Rptr. 2d 781 (Cal. App. 1993) (holding that “as is” and assumption of risk clauses in ski rental agreement did not bar recovery for skiing injuries caused by defective ski); Wheelock v. Sport Kites, Inc., 839 F. Supp. 730 (D. Haw. 1993) (holding that release agreement barring gross negligence claims against manufacturer and provider of paraglider was void as against public policy); Henningsen v. Bloomfield Motors, Inc., 161 A.2d 69 (N.J. 1960) (holding that agreement disclaiming implied warranty of merchantability was void as against public policy).

[69]See Office of New York State Attorney General Eliot Spitzer, Judge Orders Software Developer to Remove and Stop Using Deceptive and Restrictive Clauses (Jan. 17, 2003), <>.

[70]For a helpful exposition of these themes, see Donald MacKenzie, Knowing Machines: Essays on Technical Change (1996); add others.

[71]For an argument that DRM infrastructures also should be designed to preserve user privileges available under copyright law, see Dan L. Burk & Julie E. Cohen, Fair Use Infrastructure for Rights Management Systems, 15 Harv. J.L. & Tech. 41 (2001).

[72]17 U.S.C. § 1002.

[73]17 U.S.C. §1201(k)(2).

[74]See Burk & Cohen, supra note __.

[75]Cf. Cites re minimization.

[76]A desire to generate profiles of users’ intellectual preferences, for example, is privacy-related (albeit inversely) and would not count.

[77]See Cohen, supra note __.

[78]See, e.g., I am using the term “law” very generally here to encompass both legislation and regulation. A digital consumer’s bill of rights could come from Congress, but it could also come from the FTC pursuant to its mandate to regulate “unfair” practices in commerce. See supra pp. ___.

[79]See Mark A. Lemley & David McGowan, Legal Implications of Network Economic Effects, 86 Cal. L. Rev. 479 (1998).

[80]See, e.g., Berkeley clinic filing with OASIS group.

[81]Discuss OASIS vs Palladium, etc.

[82]The research that led to the development of the TCP/IP standard is a less formalized example of this dynamic. The government funded much of the initial research, and wanted to researchers address the problem of building a distributed communications network that could survive a military “hit” to some part of the network. The result was the extraordinarily robust, end-to-end architecture that we know today as the Internet. See Janet Abbate, Inventing the Internet (1999)).