Information Technology and the Law

Here’s the first change that I’ve seen due to the Grokster decision. Bonpoo is a service that lets you send large files to other people. It used to be general-purpose; you could send anything to your friends. Now, post-Grokster, they only let you send photos:

IMPORTANT NOTICE:
At bonpoo we are constantly testing file transfers services that help people send legal files across the Internet. Given the recent Supreme Court decision we have suspended our free file transfer services except for photos. We apologize for any incovience. Please check out our professional product HeavyMail for an alternative to our prior service.

Grokster didn’t win, but I think the Supreme Court preserved one of the most important ideas from Sony: the technology itself is not illegal, it’s the way it’s promoted and used that’s illegal.

I thought the structure of the decision was interesting. We have a unanimous decision written by Justice Souter that completely ignores the Sony test. Then, there are two concurring (but totally opposite) opinions by Justice Ginsburg and Justice Breyer arguing about the correct interpretation of Sony. Fundamentally, Ginsburg and Breyer read the very text of Sony in different ways.

• Ginsburg implies, “The time-shifting and library-building(?) uses of the Betamax were found to be fair use; therefore, a significant number of [potential uses of the Betamax] were non-infringing, so Sony was not liable.” In Grokster, we have “no finding of any fair use and little beyond anecdotal evidence of non-infringing uses.” Ginsburg calls on Grokster to show that “a reasonable prospect that substantial or commercial non-infringing uses were likely to develop over time.”

  Ginsburg seems to completely eviscerate the “capable” part of the “capable of non-infringing uses” test. She seems to imply that if time-shifting and library-building were not fair use, then Sony would have been liable. After all, there is no way that Sony could have predicted the billion-dollar market success of pre-recorded home videos as a reasonable prospect of a substantial non-infringing use.

• Justice Breyer, on the other hand, reads the Sony decision as formulating the “capable of non-infringing uses” test and then applying it. Even if library-building and time-shifting were not fair use (and given that only 9% of VCR use was “authorized”), Sony would still not have been liable. Breyer likes the “capable of non-infringing uses” test because it’s forward-looking and doesn’t prematurely cut off technology.

These are totally different readings of Sony. I don’t think there’s any way to reconcile them. We could have guessed at Ginsburg’s reading when she made this objection in oral argument:

There is a statement — one could take it as clear — “capable of substantial noninfringing use.” That would be very clear, I agree. But Sony goes on for 13 more pages. If the standard were all that clear, it would have stopped there. … Or if you then read back, as a careful reader would, then you find the statement that the primary use of the Sony machine for most owners was time-shifting, a use that the Court found either authorized or fair, and, hence, noninfringing.

Well, that’s the last word. How will our world change now?

I just wanted to point out that the Miller test uses “community standards” only to decide whether material appeals to the prurient interest. The second and third prongs of the test do not rely on community standards, and in fact, the second prong requires the narrow definitions of specific obscene acts that Eric and Jeff want.

The second prong of the Miller test says that material is obscene only if it describes/depicts sexual conduct specifically defined by applicable state law; that is, the Miller test requires state law to specifically define what sexual conduct is obscene. For example, here is the New York State definition of obscenity:

Any material or performance is “obscene” if (a) the average person, applying contemporary community standards, would find that considered as a whole, its predominant appeal is to the prurient interest in sex, and (b) it depicts or describes in a patently offensive manner, actual or simulated: sexual intercourse, sodomy, sexual bestiality, masturbation, sadism, masochism, excretion or lewd exhibition of the genitals, and ( c) considered as a whole, it lacks serious literary, artistic, political, and scientific value.

These are exactly the narrow definitions of specific obscene acts that Eric and Jeff want.

I agree that rights that depend on fickle community standards are not rights at all. However, narrow definition of specific acts is not the way to fix it, since these narrow definitions already exist. Obscenity should receive First Amendment protection.

I would like to consider Felten’s revised Communications Decency Act (CDA-2) that uses tagging systems like PICS to identify indecent material. This CDA-2 could include an affirmative defense for website owners who label their websites as indecent, with language like:

It is a defense to a prosecution under this act that a person has tagged a website as indecent.

Restricting ourselves to just the Web for the moment, this system would allow concerned parents and children to configure their browsers to block indecent materials. Websites could comply at fairly low cost, and the threat of prosecution would seem to encourage indecent websites to label themselves accurately.

Superficially, this seems to skirt the constitutional problems. The CDA runs into trouble because it is a content-based restriction on speech; it is not narrowly tailored because of its vagueness and impact on protected adult speech, so the CDA cannot survive strict scrutiny. On the other hand, our CDA-2 has no effect on speech, and it is really just a truth-in-labeling law. So, the CDA-2 should pass First Amendment muster.

However, I don’t think this scheme would effectively prevent minors from accessing indecent material. If history is any guide, this would fail just as the V-chip failed for indecent television. (Haven’t heard of the V-chip? Exactly.) Parents won’t know that the technology exists, and kids will be good at circumventing it. While the threat of prosecution in CDA-2 will perhaps keep the tagging more accurate (or at least more conservative), we shouldn’t deceive ourselves that any voluntary filter can keep kids from accessing indecent material.

The other flaw is that this tagging scheme suffers from the undefined term “indecent,” just like the original CDA. The ACLU argued that the CDA’s vagueness violates the due process clause of the Fifth Amendment as well as the free speech guarantees of the First Amendment, but the Supreme Court accepted the ACLU’s free speech argument without considering the due process argument. While our CDA-2 escapes a First Amendment challenge, its vagueness would undoubtedly be challenged on Fifth Amendment grounds. This vagueness gives CDA-2 a chilling power on free speech despite its First Amendment constitutionality. Speech that could be considered indecent in the absence of any definition (e.g., contraception, STDs, abortion) yet very useful for teens would likely be labeled as indecent because of a fear of prosecution under this vague statute. Moreover, the risk of criminal prosecution would strongly encourage webmasters to err on the side of caution and label any questionable material as “indecent.” Thus, the ambiguity in the statute and the high cost of error would yield real chilling effects.

On a deeper level, these chilling effects expose important questions about the relationship between minors and the First Amendment, and whether we should permit more restriction on speech with minors than speech between adults. Historically, we have not given as much importance to speech with minors, recognizing a compelling governmental interest thath “children be both safeguarded from abuses and given opportunities for growth into free and independent well-developed men and citizens” (Prince v. Massachusetts). In particular, we have recognized “parents’ claim to authority in their own household to direct the rearing of their children,” and restricting speech to minors is supposed to serve that end. Is this really legitimate? If we recognize the power of free speech to upset and thereby improve our society as adults, can we justify restricting free speech for minors? If we recognize that parents should be able to direct the rearing of their children, should we forbid proselytizing to minors? We should consider these questions thoroughly before automatically assuming that First Amendment protection shouldn’t apply to minors in full and before seeking legislation like the CDA.

We speculated in class about how the billion-dollar pornography industry can exist in light of Section 1465. According to this analysis, Section 1465 is pretty much just “dead letter,” unconstitutional, and unenforceable:

Section 1465 of title 18, United States Code, [reads]

Whoever knowingly transports or travels in, or uses a facility or means of, interstate or foreign commerce or an interactive computer service (as defined in section 230(e)(2) of the Communications Act of 1934) in or affecting such commerce for the purpose of sale or distribution of any obscene, lewd, lascivious, or filthy book, pamphlet, picture, film, paper, letter, writing, print, silhouette, drawing, figure, image, cast, phonograph recording, electrical transcription or other article capable of producing sound or any other matter of indecent or immoral character, shall be fined under this title or imprisoned not more than five years, or both.

The transportation as aforesaid of two or more copies of any publication or two or more of any article of the character described above, or a combined total of five such publications and articles, shall create a presumption that such publications or articles are intended for sale or distribution, but such presumption shall be rebuttable.

First of all, I must say that all of this is blatantly unconstitutional, except as narrowly applied to “obscene” material. Many supreme court rulings make this abundantly clear. Apparently this entire law (Title 18 U.S. Code Sections 1464 and 1465), except as narrowly applied to obscene material, is considered “dead letter”. That is to say, it hasn’t been specifically struck down by the Supreme Court, but it is widely known to be unconstitutional and so it is not enforced.

Ignoring for a moment the dubious deals that Microsoft made with Apple, Intel, OEMs, ICPs, and ISVs, I would like to question the wisdom of holding Microsoft liable for the business and technical decision to tie Internet Explorer with Windows. The appellate court found per se tests inappropriate on this question and remanded the question back to the district court for a resolution based on “rule of reason.” On remand, the Department of Justice would need to show that the tying arrangement unreasonably restrained trade.

Bundling two products together can create efficiencies, and operating systems in particular benefit from this bundling. The core of an operating system is just its basic kernel — but nobody would want an operating system that included nothing else with the kernel. Most of the competitive advantages of an operating system are the tools and APIs that are layered on top of the kernel. We expect any UNIX operating system to bundle the standard UNIX tools — sh, init, ls, login, cp, mv — and I would be hard-pressed to say an operating system without these tools is a UNIX at all. Similarly, if Microsoft sold the Windows kernel — KERNEL32.DLL — without the Windows installer, graphical shell, the start menu, the desktop, and the file explorer, nobody would buy it. An operating system by its nature simply is a bundle of disparate bits of software designed to work well together; thus, the integration and bundling constitute a large part of the operating system’s value to consumers.

Is Internet Explorer fundamentally different from these other operating systems components? Arguably, there is a separate market for web browsers, while there is no separate market utilities like “ls.” However, there is a significant market for many other components shipped with a modern operating system. There is a market for Windows shell replacements (WindowBlinds, Aston) and Windows file managers. More substantially, there is an enormous market for word processors; yet, Microsoft bundles Notepad and WordPad with Windows for free, and every self-respecting Linux distribution bundles emacs and LaTeX. Do these bundles constitute an unreasonable restraint of trade?

Should we demand that Microsoft sell discounted versions of Windows without all these tools? Or that each of them be individually removable? What if another prominent OS vendor bundles some 7800 different software packages with its operating system, for free?

Tying so often yields benefits to the consumer in the OS market that it seems unfair to single out web browsing as the only component of an operating system that must be removable or unbundled. Moreover, building web browsing into an operating system really does have concrete, identifiable benefits to the consumer. Ignoring the ethics of Microsoft’s shady dealings with OEMs, bundling a web browser with Windows really is one of the most efficient ways to get a browser on every computer in the world. Should we really bar these sorts of market efficiencies?

(It may well be the case that anti-trust law under the “rule of reason” does not in fact prohibit the tying of Internet Explorer and Windows, but I think it is definitely a substantial deficiency in the law that it is impossible to know whether it does or not.)

It’s interesting that i2hub even has a Terms of Service, given Grokster’s attempt to relinquish any and all control over users. In addition to a spiel about not posting child pornography or commiting copyright infringement, there is actually a section about what i2hub will do about copyright infringement:

i2hub will terminate the accounts of users who violate copyrights or other intellectual property rights of third parties if we receive “actual knowledge” of their infringing activities. “Actual knowledge” means a ruling from a judge stating that the user has infringed copyrights or other intellectual property rights. Those users deemed “repeat offenders” by i2hub will have their i2hub account permanently cancelled. Our repeat offender policy is as follows:

In the post-Aimster and post-Grokster era, it’s interesting to note that i2hub appears to be technically designed to be unable to monitor the transfers that users make. Nevertheless, they are not willfully blind as Posner found Aimster to be. They are aware that copyright infringement can and does occur on their network; so, they will terminate users accounts if they are provided with “actual knowledge” of infringement, though their standard of “actual knowledge” is quite high.

Is this effective? Or is it just a safe harbor? Certainly, by the time a judge decides a user is infringing copyright and i2hub receives notice, i2hub will not have “actual knowledge” of infringement at a time when i2hub can actually do anything to stop the infringement. I2hub may be able to make some attempt to stop future infringement by the same person, but perhaps not even that.

From a jurisdictional point of view, the agreement is governed by the laws of Nevis:

This Legal and Terms of Service as well as all disputes arising out of or in connection with this terms shall be governed by the laws of Nevis, without regard to or application of choice of law rules or principles.

Any dispute arising out of or in connection with this Legal and Terms of Service, or in future agreements resulting there from, shall be exclusively resolved before the competent court in Nevis.

Nevis is an island in the Carribean Antilles, just west of Antigua. (Yes, really.) Now is a contract in Nevis binding for an American? Say I violate the TOS; i2hub sues in Nevis. I don’t show up to court, so they file default judgment against me. Can they actually collect?

I’ve posted the remainder of the TOS below for the curious.
(more…)

Internet2 is not nearly as closed as Sebastian suggests. It is nominally a closed research network, but members can decide to route standard Internet traffic over their Internet2 connection. In practice, many universities choose to do this, which makes any file transfers between these universities very fast, regardless of what software students use. If a Kazaa user at Princeton downloads from a Kazaa user at MIT, the transfer will use Internet2 and be very fast. The advantage of i2hub, then, is that it only permits students on Internet2 connections to connect, so you’re guaranteed that all the transfers will be between Internet2 sites and therefore fast. Plus, users get the illusion that it’s a private network and safe from prying eyes…

Anyway, it is not clear to me that any contracts or laws are being violated if an Internet2 member allows the RIAA to access the network. If your friend came to visit, and you allowed him to use your computer to check his email and access the Internet, is that illegal? The RIAA could have done much the same thing. Or, what if the RIAA asked a member institution to simply get a list of users and files from i2hub? Why would that be illegal?

Also, the membership list for Internet2 indicates that Warner Brothers is a member. Surely Warner Brothers could have monitored i2hub and given the RIAA some data?

In any case, with some 200 university members, it’s hardly a “closed” private network anymore… anybody at any of these Internet2 institutions could have gotten evidence of filesharing.