Information Technology and the Law

Majority opinion by Justice Souter; concurring opinion by Ginsberg (joined by Rehnquist and Kennedy); concurring opinion by Breyer (joined by Stevens and O’Conner).

Link.

Over the course of the semester, we have analyzed a wide range of cases that span from copyright to internet governance to free speech. Time has limited us to just these cases, but obviously, the conflicts that arise as a result of incumbent laws being inadaptable to innovative digital technologies do not end there. One issue that we did not cover that I find particularly compelling is the question of whether bloggers should be afforded the same Constitutional rights as traditional journalists. This controversy was raised in the recent lawsuits filed in the California Superior Court, Apple v. Does and Apple v. dePlume.

In Apple v. Does, Apple subpoenaed the ISPs of two online enthusiast websites Apple Insider and PowerPage, to obtain information about who leaked confidential information about products to these well-reputed news sites. The Court did not address the Constitutional issues related to this case, but decided instead that discovery should proceed because trade secrets are involved. This case is currently on appeal.

The First Amendment guarantee of the freedom of the press is pillared on the right of journalists to protect their confidential sources. This guarantee promotes the free flow of information and maximal discourse without fear of prosecution. The traditional definition of “journalist” is now under scrutiny with the recent rise of bloggers as online journalists. The blogosphere has dramatically changed the landscape of information dissemination in the online medium. Popular blogs, such as Gizmodo and DailyKos, are read by thousands of people a day and offer many unique and informative filings not covered by traditional media sources. This democratization of the tools necessary to distribute information widely to the public is undeniably beneficial to society.

But, what defines a journalist in the blogosphere? EFF describes the test as

…whether the person seeking to invoke the reporter’s privilege had the intent to use material-sought, gathered or received-to disseminate information to the public and whether such intent existed at the inception of the newsgathering process. Under this test, courts have provided the privilege to non-traditional journalists, including book authors and documentary filmmakers.

To be sure, blogs run the gamut in terms of quality of reporting. Some blogs certainly look and act like highly edited, accountable traditional media sources. Some bloggers have even obtained traditional press credentials, such as at this year’s Republican and Democratic National Convention. At the same time, many blogs often display very poor editing, low integrity, and possibly even misleading information. Where should the law draw the line in defining “journalists” and whose journalistic rights will be protected? This question has not yet been answered by the courts.

Many more important issues are also at stake including privacy and surveillance with RFID technology, and secure and fair elections with electronic voting machines, just to name a few. As currently unrealized technologies make the leap into the public eye, existing laws will be held under microscopic scrutiny, re-interpreted by the Courts and re-written by legislatures. The controversies and debates will only get more exciting from here.

Skip ahead to September 2006. The Memorandum of Understanding between ICANN and the U.S. Department of Commerce is about to expire. Let’s make the (possibly crippling) assumption that both parties have expressed an interest to make major institutional changes to the current system. What should be done to legitimize ICANN’s position as a “global consensus entity” and policymaker for DNS? The Internet Governance Project has laid out the three possibilities:

• Top-down accountability with international oversight
• Bottom-up accountability with democratized legitimacy
• Peer-to-peer accountability with coordinated competition from the UN’s International Telecommunications Union (ITU)

The bottom-up scheme is cogent in theory, but will be virtually impossible to implement. One option calls for global direct elections. This might have worked in the early days of the Internet, but direct elections are now impractical with the number of Internet users in the order of millions, not hundreds. Furthermore, holding secure online elections with a constituency that is ill-defined will prove to be technically infeasible. Alternatively, accountability could be based on stakeholder groups representing specific social, not political, interests. A solution of this form will only further undermine ICANN’s (il)legitimacy as under-represented voices will question ICANN’s arbitrary exclusion process.

A peer-to-peer model with ICANN sharing the DNS monopoly with the ITU only shifts, and doesn’t fix, ICANN’s underlying legitimacy problem. This proposal partitions the governance of each top-level (both generic and country code) domain into one of the two regulatory agencies. Individual domain registrants will still have no choice of which agency to use. Skeptics will still question ICANN’s legitimacy to govern the subset of TLDs for which it is assigned, and this model does not address ICANN’s transparency and accountability. Moreover, there is no guarantee that the ITU will actually offer more desirable policies than ICANN and certainly will not provide overarching policies that can be consistent with varying national laws. ICANN and ITU wouldn’t be checking each other’s power in a market-like system; rather, they would be two distinct power centers.

The top-down approach with accountability to an international organization, such as the United Nations, is the most equitable and realistic solution. This strategy would delegate specific trademark dispute responsibilities to existing governments, many of which already have judicial infrastructure to handle exactly these cases. By delegating control along political boundaries, specific policies could be tailored to the laws of each individual nation. For example, data protection laws in Europe are inconsistent with ICANN’s policy to publish all domain registrants information in the public WHOIS database. By allowing each country to govern its own ccTLD, fine-grained laws can be applied in specific locales instead of global policies.

Furthermore, top-down accountability will rebuke the three major criticisms, as summarized by the Duke Law & Technology Review, that ICANN suffers from today.

• First, ICANN will return to managing just the technical aspects of DNS, while handing down policy decisions (i.e. accredited domain registrars and dispute resolution) to existing governments. The UN, with ratification from an international majority, would make larger decisions such as the addition of new TLDs to the root DNS system.
• Second, ICANN’s (now purely technical) responsibilities will have legislative oversight by the UN, erasing the major questions about ICANN’s lack of consensus when setting policies.
• Third, concerns about the legality of ICANN as a California non-profit corporation under the auspice of the U.S. Department of Commerce will no longer be of issue.

Policies being set at the individual government level will be accountable to the public through traditional government facilities, not ICANN. A potential pitfall of this structure is its dependence on existing governments that are democratic and accountable to its constituency. Perhaps the UN should set an internationally-recognized “default” policy for nations incapable of regulating its own ccTLD. Even still, a top-down accountability model would be a major improvement on the current ICANN system, and a more realistic fit than the two alternative models.

The Robots Exclusion Standard is a voluntary protocol that defines the access privileges of robotic crawlers visiting a website. Websites specify privileges in a file called robots.txt and well-behaved robots cooperate by abiding to the requested access provisions. The problem is that not all bots are well-behaved. This standard was set not by any governmental or official standards body, but rather by consensus in 1994 by robot-enthusiasts on the “robots mailing list.” Given this history, I argue that a robots.txt file by itself should not be considered a sufficiently clear notice of access rights and its existence should not constitute a legal contract that binds all visiting bots.

Though the standard is respected by the majority of today’s robots, none are legally required to follow this ad hoc standard. It would be difficult to assert that posting the robots.txt file shown below would constitute a legal contract with implied consent by a visiting bot.


User-agent: *
Disallow: /


This code is simply syntactic notation that has no well-understood semantic meaning for those who do not wish to comply with the standard. No one is required by law to learn how this standard works or to program their bots to look at, much less respect the contents of, this file.

Furthermore, legal enforceability of such a contract would be difficult without a reasonable definition of what a robot actually is. In both the eBay and Register.com cases, the main trespass to chattels finding hinged on the fact that allowing one fast bot would encourage other bots to similarly overload the server, leading to potential irreparable harm. But what if I deploy a slower program that only queries the site once every five seconds, and could show that even an impossibly large number of similar bots would not overload system resources? This would be automatic, but would my program be considered a robot? Alternatively, what if I hire someone whose sole job is to manually traverse and cache websites (provided this guy is really good at what he does, essentially automatic) with effect identical to the slow program? Would he be forced to respect the robots.txt file too? It’d be impossible to draw a clean line delineating robots from non-bots, and thus impossible to implicitly enforce a contract on an indefinite set of online principals.

If such a Robot Exclusion Standard could somehow be enforced by law, then it seems that I should be able to define my own syntactic standard for legally excluding anyone, perhaps individuals, from accessing my site. The syntax specifying exclusion might even be purposefully convoluted, yet it would serve as a legally binding contract with all individuals who access my site. Normatively, this obviously wouldn’t be beneficial for the Internet nor for society at large. Allowing legal recourse based on exclusion files might foster a permission culture that would hamper the open exchange of information on the Internet. This automatic over-propertization of websites would significantly diminish the value of the network based on Michael Heller’s idea of the “anticommons.”

Based on my above arguments, I could go further to suggest that Terms of Use posted on websites should also not be enforceable contracts without some form of out-of-band notice. For example, I start a company called Boogle (trademark infringement aside) that chooses not to practice good netiquette and ignores the Robots Exclusion Standard. My bot would mechanically traverse websites and wouldn’t be able to understand human-readable Terms of Use pages posted on each site. One could argue in court that it would be unreasonable to conclude that a valid contract was formed with an unconscionable robot– there is no “meeting of the minds” involved. In order for a contract to be formed, there must exist some offline notice to the robot implementor by the website owner, that such a Terms of Use exists and that the robot is prohibited. This was the case in both the eBay (telephone request) and Register.com (cease-and-desist order) cases.

Thus, Terms of Use on websites should not be a valid contract upon initial contact with the site, but only upon further trespass after some explicit notice has taken place. This would place a higher analog hurdle on those who look to claim trespass to chattels in Web-based situations. If this is the case and direct notice really is required, posting a Terms of Use might even be unnecessary since it’s incapable of being a standalone legal mechanism anyway.

Stanford law professor Lawrence Lessig and Wilco lead singer Jeff Tweedy spoke last night to a sold-out New York Public Library audience about culture and creativity in the age of the Internet and peer-to-peer technologies. The discussion, titled “Who Owns Culture?,” was moderated by WIRED contributing editor Steven Johnson. As you might know, Tweedy’s band Wilco produced an album called Yankee Hotel Foxtrot in 2001 that was rejected as a failure by their record label, Reprise. The band decided to purchase the rights to their tracks back from Reprise and stream them for free off of their website. Their popularity skyrocketed on the Internet, their album was picked up by another label Nonesuch Records, and it eventually sold more than a half-million copies.

Unfortunately, the recording industry has been waging war on the same “pirate” technologies that made Wilco a success. Tweedy believes that the recording industry, with their intimidating lawsuits, fosters a feeling that listening to downloaded music is a crime. Lessig is even more disturbed that a whole generation of kids are being raised with contempt for the law– in their eyes, they disrespect the law when such ordinary, innocent actions are deemed criminal. Their discussion never addressed Grokster specifically, but a similar argument was made in the industry’s brief about “a culture of contempt for intellectual property” with an entirely different conclusion: shut down the technology. The RIAA does have a moral right to sue direct infringers, but rejecting the technology itself irreparably harms society’s ability “to promote the Progress of Science and useful Arts” in Lessig’s macroscopic view of creativity. It is time for the industry to understand and adapt to these new technological changes, he says.

This reminds me of two recent blog entries by Mark Cuban, co-founder of Broadcast.com, who contends that “RIAA sales are down because they lost marketshare. There are more CDs being self published or released by non-RIAA members than ever before.” Cuban also predicts the near-future extinction of the CD, supplanted by an explosion of sales of digital music. These two events, made possible by new democratizing technologies, severely threatens the current business model of the RIAA; but as Cuban asserts, the RIAA does not represent all artists and creators– far from it. There are undoubtedly many more independent artists now empowered with the ability to distribute content without the big record labels. As Tweedy admits, Wilco could distribute on its own and doesn’t really need a record company at all. He would rather have people hear his music and dislike it than people not hear it at all because they cannot afford it.

Lessig also expressed his usual dismay at the overturning of the public domain tradition in the United States (see his latest book, Free Culture: How Big Media Uses Technology and the Law to Lock Down Culture, for further details.) The Constitution stipulates that copyrights should only be reserved “for limited Times,” but Congress grants copyright extensions so regularly that copyrighted works are no longer passing into the public domain. As of now, not a single copyrighted work will join the public domain until 2019, and surely before then, Disney will lobby Congress for yet another term extension. Lessig points out that all this is happening in a critical age where creativity has been democratized by digital technologies, where anyone can easily re-mix digital media into highly creative derivative works at low cost. The quintessential example here is the award-winning documentary, Tarnation, which was produced on an Apple laptop using iMovie for just $218.

Tweedy, as a voice for artists, believes this current climate of copyright lockdown and legal battles makes art increasingly hard to produce. Even as a non-techie (”I’ve never seen a Powerpoint presentation before [today], but [Lessig’s] was fantastic!”), Tweedy realizes that getting rid of P2P will eliminate access to tons of previously inaccessible material, copyrighted or not. The recording industry, by locking out access and fair use, is effectively asking artists like Tweedy to invent and create new works of art in a cultural vacuum– an impossible task. Artistic works, he says, inherently depends on past culture, and this culture must be available to inspire truly great art in the future.

The doctrine of fair use has allowed for this incremental layering of culture on top of past culture, but Lessig notes an interesting problem: let’s say you make a derivative work using someone else’s copyrighted work. If the market hates you and you never become popular, then your derivative work is considered a fair use. On the other hand, if the market embraces your work and you do become popular, then your work might be considered an infringing use if sued, even though your intent and use was identical to the first situation. Lessig goes as far as to say that fair use is “the right to hire a lawyer to defend your right to create.” This discussion about fair use seems to draw parallels to one of the main questions in the Grokster case. That is, how will inventors of new technologies know a priori at the time of invention that they will not be held liable “out of the box” for secondary copyright infringement? In a similar fashion, how can artists today know at the time of manifestation that their derivative work of art won’t be sued for copyright infringement? Perhaps we need to establish a bright-line rule for fair use, similar to the Betamax doctrine for technological innovation, to clearly delineate the boundaries of creativity.

In all, Lessig asserts that we need to hear more from artists like Tweedy in the copyright debate. These creators are the ones that the RIAA doesn’t want us to see, but ones that society will surely see more of in the future.

[On that note, Stephanie Wang, a first-year Economics graduate student, and I are planning on starting a FreeCulture.org student group here at Princeton. FreeCulture.org, inspired by Lessig’s work, is a national student movement with the goal of defending innovative technologies and the freedom to create. We strive to spark discussion about these increasingly important issues at the intersection of public policy and technology, and provide an appropriate forum for these discussions and debates to take place on campus. Since this will be an undergraduate club, I’d love to hear from you if you are interested in getting involved, especially if you are not graduating this year! E-mail me at harlanyu@.]

The Grokster oral arguments on Tuesday morning gave us a glimpse into the minds of the Supreme Court Justices. The questions they asked showed that they clearly understood the big picture and the broad implications this case has on the future of innovation. This understanding is undoubtedly good for Grokster and the Justices’ questions made me quite optimistic as to the outcome of the case. Below, I’ll attempt to break down what each Justice is thinking based on the oral arguments. This analysis may give us a hint about the nature of the Court’s decision that will be released sometime in June or July.

As a disclaimer, I had an obscured view from the very last row of the Courtroom. It was sometimes difficult to tell which Justice was speaking from where I was situated and I could not see Justices Ginsburg or Souter at all. I’m using my own notes and my limited memory, which might possibly attribute statements with the wrong Justice, but I have tried to confirm each Justices’ views and questions using external reports of the same arguments.

In order of seniority:

Despite his health problems, Chief Justice Rehnquist presided over the case and asked only a few short clarifying questions. I couldn’t clearly decipher much of what he asked mainly due to his very raspy speech, a result of his throat surgery from thyroid cancer. He also momentarily left the Courtroom during the arguments on a number of occasions with the help of aides. Note that Rehnquist was a dissenter in the 5-4 Sony Betamax decision in 1984, which may put him loosely on the side of the entertainment industry.

Donald B. Verrilli Jr., the attorney representing MGM et al., presented the figure of 2.6 billion infringing files downloaded per month that represents 90% of all downloaded files. Justice Stevens recognized that this still amounts to millions of actual legitimate uses on peer-to-peer file sharing systems every day. I don’t recall him being very active in questioning Richard Taranto, the Grokster and Streamcast attorney, so I would expect the author of the 1984 Sony Betamax majority decision to lean towards protecting innovation and legitimate uses.

Justice O’Conner was the swing vote for the majority in the Sony Betamax ruling. During Tuesday’s arguments, she was very interested in getting a clear answer on what standard the industry was proposing. She asked many questions about the active inducement test, and grilled Acting Solicitor General Paul Clement on his interpretation of Sony. She wanted to know why Clement thought the 9th Circuit got it wrong, and Clement responded that there was only anecdotal evidence of non-infringing uses. Clement then surprisingly continued to suggest applying a majority use test, further clouding any clear standard proposed by the Petitioners. Based on Verrilli and Clement’s apparent confusion of what standard to use (adhering to “substantial non-infringing uses”, using the IEEE active inducement standard, or applying a majority use test), Justice O’Conner probably came away dissatisfied with the answers she got. In questioning Taranto, she wondered whether or not other past bad acts by Grokster will still be available for litigation in the lower courts after this trial. Taranto answered yes; this opens up the possibility that O’Conner might intend to establish a clarified rule generally, but in specific, leave further litigation of Grokster to the lower courts. It’s likely that Justice O’Conner will be instrumental in redefining the meaning of the Betamax precedent in the decision, or will even elect to establish an updated test for secondary liability.

Justice Scalia was one of the most active questioners during the session and in my opinion, he asked the toughest and most pressing questions. He was particularly uneasy about the effects on future innovation, asking Clement how many years a new technology should be given to mature before evaluating whether or not its innovators would be liable. Moreover, he wondered how an innovator would know at the time of invention whether or not he would be sued “out of the box.” He seemed unsatisfied with Verrilli’s claim that the iPod was a legal foregone conclusion with inherent non-infringing uses at the time of invention, while a technology such as Grokster was not. At the same time, he was disturbed by the issue of willful ignorance and commented that Grokster might have disabled the capability of obtaining specific knowledge solely to get around the Napster ruling. He also stated outright that the Court will not decide this case on stare decisis such that arguing “Grokster is just like Betamax” will not go very far for Respondants and that the Court will decide this case on independent grounds.

Justice Kennedy openly deplored the entertainment industry that it was unclear what test they were proposing from their submitted briefs. He was very concerned with the fact that Grokster relied on illegal activity as start-up capital for their business, and said that their business model “just seems wrong to me.” Taranto’s argument about looking only at current business practices and separating out past bad acts didn’t seem to satisfy him. Kennedy’s pointed questions about Grokster’s business model leads me to believe he won’t fall on the side of Grokster, but also will not easily accept the industry’s scattered arguments.

Justice Souter was sympathetic to “the guy in the garage” and like Justice Scalia, he wanted to know how inventors could be confident in advance that they wouldn’t subject to liability under the Petitioner’s standard. Grounded in reality, Justice Souter seemed confused as to why Verrilli thought the iPod was unquestionably legal but other technologies were not– and where this line should be drawn. He adeptly recognized many infringing uses of the iPod, that if music could be acquired for free, people would do just that to fill their iPods. On that other hand, he hypothetically applied in the active inducement standard in questioning prior bad acts, saying that inducing customers on Monday through Thursday would effect the sales of a product sold on Friday. He seemed unsettled about the potential of willful ignorance by Grokster and only received a narrow answer from Taranto (that willful ignorance implies having specific information, which Grokster did not, and ignoring this information.) Justice Souter was also involved in trying to decipher a clearer standard and wondered if Petitioners were trying to impose some sort of “flexible rightness” doctrine.

Apparently, Justice Thomas never asks any questions in court, and this day was no different. Often lounging back in his chair and looking generally unengaged, it’s impossible to tell what he is thinking. I could not find any external information about his general stance on technology, but I would assume he is relatively well-versed in understanding technological issues as the youngest Justice on the Bench.

Justice Ginsburg was irked by Taranto’s suggestion that Sony is a bright line rule that has protected innovation and should not be altered. She claimed that the Sony rule was really not all that clear, as the Court writing the majority decision would have stopped after the rule but instead continued on for 13 more pages. Like O’Conner, she seemed primarily interested in finding the proper balance by clarifying the standard set out by Sony. She also worried about giving technologies a free pass and an economic incentive to maximize infringement, as Verrilli argued.

Justice Breyer dropped some very tough questions on Verrilli, asking him how peer-to-peer technologies were really any different from the iPod, the Gutenberg press and the Xerox copy machine. He claimed that these past technologies were undoubtedly good for society, and that the Betamax standard has served innovation well for the past 21 years. He also understood that technologies, such as the Xerox, had many foreseeable infringing uses at the time of invention and that the industry’s proposed standard might stunt innovators lest they be sued. In addition, Justice Breyer asked Taranto about his interpretation of the Sony precedent and the difference between the system’s capabilities and its actual uses.

In all, Justices Scalia, Souter and Breyer were all squarely concerned about the impact on new inventors and future innovation, while Justices O’Conner, Ginsburg and Kennedy were most interested in the interpretation of Sony and the merits of the other proposed tests. It was difficult to glean the stance of Justices Rehnquist, Stevens and Thomas from the oral arguments alone.

My best guess is that the Court will not issue a decision that drastically hurts future technological innovation in general. It was comforting to hear that more than a few Justices recognized possible problems that future innovators would face if they reverse the 9th Circuit based on the existing Betamax doctrine. I believe that the Court will attempt to clarify the existing Betamax doctrine or will adopt a new test altogether that incorporates some conception of intent or inducement. I also presume that the Court will not issue a decision that gives companies, with Grokster-like business models, a free pass to induce infringement. The Grokster decision might very well turn out to be the landmark decision for the upcoming era of technological innovation and I am relieved that the Justices seem to fully grasp the issues important from an innovator’s point of view.

The evolution of software distribution has been a unique one. Unlike the purchase and subsequent ownership of a book, purchasing software only grants one a license to use the product while the title of the software remains with the creators. In this environment, software users are not only subject to the restrictions (and protections) of federal copyright law but also to the arbitrary contractual terms of the end user license agreement (EULA). Users almost always blindly click through these EULAs and agree to a long and complicated set of terms without first reading it, tilting the copyright balance in favor of the copyright holders.

This is the landscape today, but should there really be a qualitative difference between the purchase of a book versus software? In the past, the courts have rejected licensing schemes attempting to expand the copyright protection for books and phonographic recordings. Why do the courts now acknowledge the legitimacy of software contracts and enforce their provisions? One reason seems to stem from the inapplicability of the first sale doctrine in digital contexts. In Davidson v. Internet Gateway (aka “bnetd,” 2004), the Court holds that

Under the first sale doctrine, “a sale of a lawfully made copy terminates a copyright holder’s authority to interfere with subsequent sales or distribution of that particular copy.” Adobe Sys. Inc., 84 F.Supp.2d at 1089 (citations omitted). “The first sale doctrine is only triggered by an actual sale.”

The Court found that since software is never sold, the first sale doctrine is inapplicable to licensed software. In the past, it was relatively easy to define each “particular copy” as a tangible object and regulate the transfer of the single object. Now, digital products are much more difficult to regulate, so software creators probably shifted toward a licensing scheme to place an additional layer of restrictions on their copyrighted works. Normatively, it seems that without licensing, the software copyright holders could be short-changed since licensees could effortlessly sell a copy of their software without deleting the original. But is the current licensing scheme the best way to re-balance copyright for software?

Software licensing swings the copyright balance too far back into the creators’ court. The creators have a strong incentive to construct their EULAs in an unnecessarily broad manner. As discussed earlier by ‘MRK’, though it is true that creators need to protect themselves from frivolous lawsuits and other liabilities, there is no deterrent constraining them from including wildly excessive terms. Given that most people blindly click through the EULAs, software creators are able to first obtain over-broad contractual agreements and later pick and choose what violations to litigate.

A standard contract also requires “a meeting of the minds”– a measure that can rarely be attributed to EULAs. Excessive EULAs are written in such convoluted legalese that their meaning can only be fully grasped by an extremely small population of legally and technically minded experts, if ever. For example, the Blizzard EULA stipulates that a licensee may not “in whole or in part, copy, photocopy, reproduce, [or] translate… the Program.” What exactly does this mean and what specific rights would one be signing away? A standard interpretation would disallow a licensee from giving a friend a burned copy of the software CD. But, a loose translation of copying and translating can also mean that it is illegal to even run the program, since the computer would be copying pieces of the software to RAM. This ambiguity of language seems to blur the “meeting of the minds.”

This is not to advocate even longer and more technically and legally obscure EULAs, but I am skeptical that software licensing on top of federal copyright law is the right solution to re-balance copyrights in the digital domain. The conflicts between contract and copyright leaves no bright line standard for what rights are granted to software users. Further questions are raised when we consider hybrid products such as E-books– is this a book or a piece of software? Perhaps we can update the first sale doctrine to embrace new digital technologies and protect software creators without utilizing licenses, but that will be a subject for another time.