Intellectual Property Protection Regimes in the Age of the Internet by Eric Goldman

Intellectual Property Protection Regimes in the Age of the Internet

by Eric Schlachter, Esq.
Cooley Godward Castro Huddleson & Tatum, Palo Alto, CA


“For copyright, the implications are fundamental. Established notions about copyright become obsolete, rooted as they are in the technology of print. The recognition of a copyright and the practice of paying royalties emerged with the printing press. With the arrival of electronic reproduction, these practices become unworkable. Electronic publishing is analogous not so much to the print shop of the eighteenth century as to word-of-mouth communication, to which copyright was never applied.” Ithiel de Sola Pool, Technologies of Freedom (1983) at 214.

“Intellectual property law cannot be patched, retrofitted, or expanded to contain the gasses of digitized expression any more than real estate law might be revised to cover the allocation of broadcasting spectrum.” (Barlow).

To some, the Net poses the death of intellectual property protection regimes as we know it. Others, such as the NII Task Force, find that “[t]he coat is getting a little tight. There is no need for a new one, but the old one needs a few alterations.” (Footnote omitted). Others have responded that “the Emperor’s clothes still fit just fine.” (Rose, Emperor’s Clothes).

Features of the Net. I make the following presumptions about the future technology applicable to the Net.

  • Unlimited, costless (or nearly costless) distribution of information. This includes the primary distribution of the information between users and correspondence between software program or information file and a home base located on the Net.
  • Users’ ability to make identical copies. I presume copy protection will not be meaningful except as noted.
  • Anonymous remailers will allow distribution without identification. I also presume the availability of digital cash or an equivalent, which will be untraceable.
  • Tamper-proof or difficult-to-tamper-with encryption mechanisms, including a digital signature.
  • Smart agents with ability to perform actions in any publicly accessible space on the Net, at low or no cost.

Culture of the Net

To craft an intellectual property regime tailored to the digital age, it is important to evaluate the context to see how users actually behave. This provides a clue to whether it is realistic to expect users to conform their conduct to existing or proposed intellectual property protection regimes.

There are multiple perspectives about intellectual property rights on the Internet (see Rose, Is Copyright Dead?). These perspectives include:

  • A. “Information Wants to be Free.” These people believe there should be no copyrights or other protections of intellectual property; everything made publicly available should be public domain.
  • B. “Right of Attribution.” These people believe that the only rights owed to authors and creators is the right of attribution; otherwise, all information is free.
  • C. “Limited Use Rights.” These people believe that copyright has validity but minor infringing behavior, whether “fair use” or not, should be legal.
  • D. “Strong IP Regimes.” These people adhere strictly to intellectual property protections.

(Note that there is also the moral rights perspective, which exists on a different scale but is most closely aligned with the Strong IP Regimes category.)

As Lance Rose has pointed out, no one segment dominates the other; rather, we can expect that each segment will continue to attract adherents well into the future. The issue then becomes whether we want to craft our laws to override the beliefs of people in categories A and B, or if we want to primarily conform the behavior of people in categories C and D.

Lance Rose has argued that we should acknowledge that there will always be people in Category A, regardless of the intellectual property protection regime (Rose, Emperor’s Clothes).

There are a couple of other Net cultural aspects worth noting:

  • “Correction of Errors.” The Net culture promotes discussion of acceptable behavior. If someone submits a violating post, the Net’s response is egalitarian and swift: the poster may be publicly flamed, privately chastised, or even added to the user’s “bozo filter” and henceforth silenced from reaching that user. This holds true in the context of intellectual property violations (Samuelson–arguing that such behavior is an extension of netiquette). An infringing posting may be subject to the Net’s rebuke, and if the venue is moderated, may not be posted at all–not because of legal liability, but because it violates the collective sensibilities. Note that as the Net population has exploded, this trait has not yet abated. Rather, it appears to have remained an integral part of the socialization process (i.e., Newbies, once flamed, realize the value of conforming to the social norms).
  • “Censorship is a malfunction.” (EFF co-founder John Gilmore, quoted in Barlow). The Net routes around barriers to the free flow of information, whether legal or technological. As Barlow asserts, the result is that information “more perfectly adapts” to its environment.

Categories of Intellectual Property

There are different types of intellectual property circulating on the Net that raise similar but distinct issues.

I draw a distinction between “software” and “content.” This is not a rigid distinction. As I conceptualize it, software can be used to create content. This is somewhat tautological. For a more rigorous analysis of the distinction, see Samuelson et al., where they describe software as possessing “behavior.”


In determining a proper protection regime for software, it is useful to assess exactly what “property” exists in software. This is not as obvious as it appears.

Samuelson et al. argue that software is more like a machine than an authored text. This suggests that copyright regimes are inherently unsuited to protect the value of software; rather, software is more like “industrial knowhow.”

As a result, Samuelson et al. argue that one possible solution is to create a “registration depository” where creators of significant advances can register innovations and receive some protection for such registered concepts.

Contrast the Samuelson et al. position with Dyson’s assertion that the software market has become so competitive that little value is created from the software functions themselves. Dyson believes that currently value from software is being generated from the distribution chain, which ultimately is likely to also lose value as Net distribution becomes costless (or nearly so).

In Dyson’s world, software will become valueless because neither the functions nor the distribution warrant value. In such a world, creators will derive value from software creations by using such creations to create demand for ancillary services (technical support, skilled technicians, systems integration, customized proprietary programming).

Note that such a model is predicated on proper attribution, or else the creator will not be recognized and therefore not solicited for the ancillary services. Attribution may be facilitated by digital signatures. The NII task force also recognized the importance of attribution, even in the context of infringing behavior, by proposing to criminalize the fraudulent removal of identifying information from a creation.

As should be clear by now, the judgement over the created value in software completes affects the choice of a proper model. In Samuelson et al.’s world, the proper response is essentially more regulation. In Dyson’s world, existing copyright models are adequate; if anything, “infringing” distribution/redistribution of software should be encouraged.

Choosing between the two models, and their offspring and derivatives, is difficult. However, as Lance Rose has noted, we have seen numerous successes of existing copyright laws of protecting the interests of creators while recognizing “normal” human behavior. As Rose’s piece is entitled, “the Emperor’s Clothes Still Fit Just Fine.”

Furthermore, we should recognize that many creators will not need to rely on ex post facto suits for infringement to recoup value. Rather, there is some technology and logistical processes, of uncertain efficacy, in the pipeline that could result in extra-judicial prospective protection of software. Some examples:

  • Technical Support
    • Limit technical support to registered users. This is the philosophy of many (all?) shareware companies offering technical support, including most prominently Netscape. This is a corollary of the principle of deriving value from ancillary services: Unregistered users can use the product for free, but will have to pay for support.
    • Limit technical help built into the software and rely instead on hard-copy documentation. Many shareware companies also offer hard-copy documentation as part of the benefits of registration. There is nothing to prevent photocopying or scanning of the documentation to defeat this decidedly low-technology approach, but it will require infringers to go through some extra steps that may not be worth the effort.
  • Usage Metering and Encryption
    • “Superdistribution” (Cox). Cox’s approach is to use a hardware component that calculates fees based on usage and then submits periodic reports to a home base. (This could also be done as a debit card approach–each use would offset credits on a card; when the credits are zero, the user would have to recharge the card). This approach’s advantages include that it would encourage free distribution of software and allow appropriate royalties to flow to creators of components. However, some serious concerns exist: the approach would impose hardware costs, misuse of the metering information could result in privacy invasions, and the hardware would be potentially subject to tampering.
    • Authorization codes. Software would require an authorization code that would permit the software to operate for a certain period of time. Authorization codes would be obtained (for a fee, of course) from a home base. While this does not impose hardware costs, the concerns about privacy and tampering exist here as well.
    • Software envelopes (Griswold). Copyrighted material would be contained in an encrypted software envelope that would periodically communicate with a home base upon use. The software would require authorization from the home base to continue functioning. If the user has not paid for the use, authorization would be denied and the software would be disabled. (This is somewhat analogous to the process used for satellite transmission feeds). The periodic communication could be counted to determine a metering charge. Again, the privacy and tampering concerns remain.
    • Centralized software. Software could be centralized in one location and users access the program remotely via the Net on a subscription or per-use basis. This is analogous to the old timeshare models. Once again, there are privacy and tampering concerns. It also remains to be seen if consumers, accustomed to walking out of a software store with a tangible product, will adapt to lacking any possessory interest in software.

Note that in the metering models, there will be a tremendous increase in the amount of information transmitted back and forth on the Net. These transmission demands will create costs that must be borne by someone–ultimately, the user.

  • Software-based controls
    • Limited functionality programs. These are frequently used in software programs today, with limits such as disabled printing and saving or “time bombs.” Chances are that these limited functionality approaches will merge with the metering approaches described above, since the Net will be an efficient way to remove the limits. As with the other proposals, these are subject to tampering.
    • Build “bugs” into the software, requiring users to get updated, bug-free versions (suggested and rejected by Dyson). This is a crude way to disable software and likely to produce ill-will. Furthermore, without other protections there is nothing to prevent the dissemination of the non-buggy versions once released.
    • Copy protection. Among other approaches, software could have a sensor that causes the software to self-destruct if user tries to copy (Barlow). While highly effective when it works, copy protection has fallen into disfavor with consumers and is rarely seen now.
  • Post hoc infringement control
    • Agents could scour the Net looking for publicly available infringing copies. This will not prevent private piracy but could be extremely effective at preventing unauthorized public distribution. Of course, as in the Cold War, we could see countermeasures by the pirates, leading to a classic arms race–the person with the mightiest agent wins!

Files (“content”)

There are two key types of information: loosely, information that creates value from being known, and information that creates value from being kept confidential. The latter group includes information that loses value over time (i.e., market information).

Without intellectual property protection, it is likely that information not for public consumption will be made available either by subscription/encryption or by patronage. Either way, the result will be stratification into the information poor and the information rich. In turn, this stratification will be highly correlated with wealth, leading to perpetuation of society classes.

Note that the stratification will not necessarily result from superior information, but from superior access to information. In “drinking from a firehose,” it is helpful to have someone controlling the spigot. The spigot will be controlled by indexers, who will reap much value from their expertise.

Many creators fear the leakage of “exclusive” information. Some have suggested that leakage should be supported, because it has the same effect as marketing and creating demand for access to the complete proprietary information (Dyson). There is an emerging trend of giving away valuable information to become recognized as the leader in the industry; once again, rewards come from the sale of ancillary services (ex: allowing people to tape Grateful Dead concerts to create demand for attending Grateful Dead concerts and purchasing Grateful Dead paraphernalia; Fenwick & West’s promotion of the Multimedia Law Primer, which is a promotion piece to induce browsers to buy the book, which in turn is a promotion piece to induce readers to hire Fenwick as attorneys). Others have argued that leakage is not an exclusively a threat on the Net, given the piracy that takes place with T-shirts, etc., but that there are limits to such piracy because someone engaging in the business of infringement will face much greater odds of being caught (Rose, The Emperor’s Clothes).

The issue boils down to whether copyright laws are intended to prevent all forms of infringing behavior. In the context of information intended to be kept secret, there are probably other more efficient mechanisms to control infringing behavior: a combination of trade secret law, contractual provisions, and perhaps “pipe-narrowing” to make infringement costly: charging high connect times, restricting searches, and permitting access only with proprietary programs that disable certain features (Rose, SysLaw).

In the context of information that benefits from being known, there will continue to be much benefit from distributing such information as an entree to ancillary services. Again, the most important feature in such contexts will be the right of attribution. Presumably, appropriation by indexers and other data pirates will not pose a problem so long as the attribution remains.

Much information designed for public consumption will be advertiser-supported. Perhaps we will finally see the long-rumored merger of advertising and information. Perhaps the intermixing of advertising and content, such as we see on TV and in magazines, will remain robust. So long as the value comes from ancillary services (whatever is being advertised), the right to attribution remains fundamental.

Copyright laws still have a place, but as they do today, they will only protect the very valuable copyrighted materials. To the extent these are not protected by technology, infringers will face steep penalties and holders will have incentives to protect their interests. This suggests that the existing regime has its place, as it currently has, without modification and recognizing the limitations.

For those copyrights that have very high value, the proposed copyright clearance process may prove valuable.

One other approach to revitalize existing copyright laws is to recognize the ability of copyright holders to impose “shrinkwrapped” online use licenses on content. These might say, “I give the right for distribution for educational and research purposes only.” Infringement of such licenses would give holders a right of action; one that would probably rarely be enforced, but would enhance the power of motivated holders.

Finally, there may be technology solutions to assist in preventing infringement. For example, Bell Labs has created a method to insert “serial numbers” for documents. This would imbed unique codes in documents, so that each and every copy could be traced to the original source, without being immediately obvious to any infringer. While this would not disable copying, it would provide disincentives if everyone knew the risk that the infringing copy could be traced back to them. (Voss). (See also Jim Warren, proposing a similar concept.).

Other Problems

Joint Efforts. The copyright scheme is not especially well-suited to assigning ownership interests when there is collective development. We have already seen this problem in multimedia, to which the copyright clearance process could provide some relief, but there remains the difficult issues arising from “message bases,” digital sampling, and product produced by groupware.

Another analogous problem arises in assigning the ownership of programs created by programs.

Jurisdiction. The global nature of the Net, cutting across jurisdiction boundaries, creates an inevitable conflict of laws. For example, in some jurisdictions, accessing a web-accessible document might be deemed making and distributing an infringing copy. If a user from a jurisdiction without such a law browses the document, there has been unintentional and (IMO) harmless infringement.

This problem is especially acute in the context of moral rights, which are not universally recognized in all jurisdictions. Could a holder of moral rights assert that someone lacks the right to browse their document? Note that Pepper and Corrizini have placed a remark on their Website that commercial sites lacked the right to link to their site. Would such a restriction be more enforceable in a moral rights jurisdiction?


John Perry Barlow, “The Economy of Ideas: Selling Wine Without Bottles,” Wired, March 1994. (available online at, in the Intellectual Property section, and at

Dan Burk, “Transborder Intellectual Property Issues on the Electronic Frontier,” 5 Stanford Journal of Law and Policy Review (1994) (Available online at

Brad Cox, “Superdistribution,” Wired, September 1994 at 89 (available online at

Esther Dyson, “Intellectual Property on the Net,” Release 1.0, December 28, 1994.

Jean Erhard, “Digital Rights,” Internet World, November/December 1994 at 78.

Gary N. Griswold, “A Method for Protecting Copyright on Networks,” Summer 1994. (Available online at

Information Infrastructure Task Force, Intellectual Property and the National Information Infrastructure (Green Paper), preliminary draft, July 1994. (available online at

Lance Rose, “What Kind of Online Copyright System Do You Want? Tell the NII Task Force Today!,” Boardwatch, September 1994 at 88.

Pamela Samuelson, “Legally Speaking: The NII Intellectual Property Report,” Communications of the ACM, December 1994 (Available online at

Deborah Reilly, “The National Information Infrastructure and Copyright: Intersections and Tensions,” Journal of Patent and Trademark Office Society, December 1994 at 903.

Lance Rose, “Is Copyright Dead on the Net?,” Wired, November 1993 at 112. (Available online at

Lance Rose and Jonathan Wallace, SysLaw (2d Ed.) 1992 at 59-60.

Lance Rose, “The Emperor’s Clothes Still Fit Just Fine,” Wired, February 1995 at 103 (available online at

Pamela Samuelson, Randall Davis, Mitchell D. Kapor, and J. H. Reichman, “A Manifesto Concerning the Legal Protection of Computer Programs,” Columbia Law Review (forthcoming).

William S. Strong, “Copyright in the New World of Electronic Publishing,” June 17, 1994 (Available online at

David Voss, “Stop That Copy,” Wired, August 1994 at 34 (available online at

Jim Warren, “GovAccess.107,” March 12, 1995.

Also, check out articles at