Why Operation In Our Sites Is Illegal And Needs To Be Fixed ASAP
from the speech-violations dept
It's worth reading the whole thing, but among the key points is how this "right" to seize domains for supposed copyright infringement is one of the things that was quietly slipped into law via PRO-IP a few years back. While this point hasn't received much attention, it is important to discuss, because it highlights how the RIAA/MPAA lobbyists have a history of sneaking little timebombs and easter eggs into new laws proposed by "friendly" lawmakers. There were so many things that people were concerned about with PRO-IP that this issue of "harmonizing" seizure laws between copyright and counterfeiting was something no one paid attention to because, honestly, no one realy thought that the US government would start censoring websites. Everyone assumed, falsely it turns out, that the intention of this part of PRO-IP was to allow for the seizure of physical tools, not domain names. This is another reason why we should be fearful of PROTECT IP. Who knows what other "goodies" the RIAA and MPAA have hidden in there that will pop out in a few years' time.
Of course, even though PRO-IP may have extended seizure laws to allow for more broad seizures in copyright cases, that still doesn't make these seizures legal or Constitutional, in any sense. Defenders of the seizures often point to Terry Hart's analysis of the seizures as legal on First Amendment principles, but Sellars does a nice job dismantling Hart's analysis and showing how the seizures do not fit with existing jurisprudence on prior restraint (footnotes removed):
And the question here is not whether substantive free speech law would save the websites from copyright liability. It is clear that it would not, if adjudicated as infringing. Instead, the question is whether the in rem seizure of pure speech, done because it may be infringing, violates the procedural safeguards instituted by the First Amendment. Here, entire websites consisting of pure expression were removed, presumably because some of the speech encouraged viewers to follow links to third-party websites and commit copyright infringement. Direct copyright infringement was not being conducted on these websites at all; it was instead the cyberlockers that reproduced and distributed the content. If the speech is adjudicated to be “inducing” under that secondary liability theory the speech is likely enjoinable consistent with the First Amendment. If it is not so found, it is constitutionally protected free speech. And at this moment no court has actually turned to the merits of this claim and determined if it is indeed unlawfully inducing. Instead, a magistrate judge determined that there was a sufficient probability that it may be infringing, and used that alone to take the website down. Courts would not tolerate such a cursory review in all other areas of free speech law.Sellars notes, as we have in the past, that the caselaw is pretty specific that seizures involving speech should be limited to areas in which the seizure is necessary to preserve evidence and isn't done solely for the purpose of blocking that speech. That is clearly not the case here.
The First Amendment embodies certain procedural safeguards to help prevent free speech from being accidentally silenced while unprotected speech is enjoined. An analogy to obscenity doctrine can provide useful guidance. Like infringing speech, obscene speech is unprotected by the First Amendment. The determination of whether a work is obscene also depends entirely upon an examination of its content, requiring application of a legal test to specific facts. And in both cases the difference between protected and unprotected speech can sometimes be a “dim line.”
Courts are quite sensitive about the precarious line between proscribable speech and protected speech, and are very hostile against government efforts that engage in “prior restraint.” Prior restraints are considered the “most serious and least tolerable infringement on First Amendment rights,” and bear a “heavy presumption” against validity. Any effort to remove speech from circulation before the speech is adjudicated as unlawful can work an unconstitutional prior restraint, even when imposed after the speech is published.
To that end, courts have recognized that seizure of expressive works requires special procedural considerations.
Needless to say, these safeguards were not followed in Operation In Our Sites. The seizures here were not done to preserve evidence, as nothing tangible was taken into custody, and it would be illogical to claim that there was any risk that a defendant would “flee” with their domain name and thus deprive the court of evidence. This seizure instead took the website out of circulation entirely, at least for a time. No administrative proceeding with the safeguards of Freedman was present. Instead, a single federal agent made a probable-cause level statement to a magistrate judge, who ruled ex parte. In striking contrast to the careful process required when seizing expressive works in the doctrine of obscenity, the seizures conducted in Operation in Our Sites show the bare minimum of process. This turns First Amendment due process on its head; it takes down speech on the basis of its content (to wit, that it is infringing content or induces others to infringe copyright) before an adversarial proceeding determines if the speech is in fact unlawful.One of the popular explanations from Hart and one particular commenter on our site is that there is no First Amendment issue here because the seizure of the domains are not because of the domain's expressive nature, but merely because they are "property" used to facilitate infringement. Sellars punctures that claim deftly:
Critics also argue that First Amendment analysis is inappropriate because the websites were not seized because of their expressive content, but because they are “property” used to facilitate crimes. This has intuitive appeal. After all, we do not use the First Amendment to stop the closure of a bar that violates liquor laws, even though bars are often places where members of the public gather to debate the issues of the day.Sellars also responds to critics who say it's important to seize these domains without a hearing or notice to stop the immediate harm of damage being done. But, as he notes, First Amendment case law on prior restraint has long held that's a meaningless argument:
The problem with this argument is it unfairly characterizes the law at issue. Copyright forfeiture is not a content-neutral law allowing for the seizure of any property used in crime. The law providing for forfeiture in copyright cases expressly incorporates substantive copyright law. Copyright itself is a content-based form of regulation: it determines the legality or illegality of speech on the basis of how the speech is expressed. To equate content based laws with content-neutral laws does not conform with First Amendment doctrine. These websites were not seized here because their domain name offended copyright. It was the speech on the website, allegedly telling people where and how to find infringing content, that was the crux of the forfeiture. The object of the domain name seizure was to constructively remove this offending speech. The analysis of illegality here begins and ends with an examination of the speech for its content.
This argument ignores the very reason why courts are so hesitant to engage in prior restraint in the first place. The same argument could be raised for defamatory speech, speech that could threaten national security, or speech that incites others to violence, and yet we do not tolerate prophylactic injunctions in those cases.Sellars also does a nice job highlighting, yet again, how these seizures don't even work, which really highlights the ridiculousness of the whole thing. ICE insists it needs to seize these domains without warning, notice or adversarial hearing to stop the infringement, and yet (contrary to ICE claims) there is no indication that these seizures have slowed the use of these websites. In many cases, the websites come back more popular than before.
Finally, Sellars raises a point that is often ignored in this: that the seizures seem to go against the intent of current copyright law itself, in the form of the DMCA. For whatever its many problems, the crux of the DMCA was to try to resolve issues of copyright infringement through the ability of rights holders to issue notices and get online services to takedown the offending content. That is, the point was to avoid any judicial or law enforcement involvement altogether. And yet, there is no indication that rights holders properly notified the seized domains of issues regarding the content that was used as the basis for the seizures. In fact, there is evidence that when many of these sites did receive takedowns, they promptly complied:
This is not just coincidence or legislative good fortune for websites that frequently link to content that may be infringing. The safe harbors of the DMCA are a legislative expression of a desire to foster cooperation between online service providers and copyright owners. The legislative history around the safe harbor indicates that the provision is “intended to promote the development of information location tools generally, and Internet directories ... in particular.” The statute mandates a bargain between these two groups: the websites will have a designated person to receive complaints and respond expeditiously to remove infringing links, and in exchange the copyright owners will not sue the websites for infringement through their linking.Of course, Sellars' paper also offers some ways to fix Operation In Our Sites, and none are too surprising or crazy. In fact, they're things that many of us have suggested in the past: require initial action be "in personam" (against those involved, rather against the site itself), meaning that there is an opportunity for an adversarial hearing prior to the seizure. To date, despite my asking many times, no one has been able to give a reasonable answer as to why the federal government does not allow for such due process. Thankfully, at least, PROTECT IP does seem to say that an attempt should be made for an in personam action before an in rem action, but there are some pretty big loopholes. Secondly, he notes that Operation In Our Sites should follow existing standards for the seizure of speech as laid out in existing case law. Specifically, he points to the standards put forth in Freedman v. Maryland. Finally, due to the clear failure of these seizures to actually stop the actions targeted, he suggests that ICE recognize the low likelihood of success and focus on efforts that might actually have a positive impact.
This spirit of cooperation vanishes when enforcement moves from an inter partes to an ex parte remedy. The copyright owner is no longer encouraged to reach out to the website owner. Instead, they will persuade ICE to file a seizure warrant to take down the disputed websites. The website owner now has no chance to expeditiously right the perceived wrong. Instead she must come forward in court and dispute the criminality of the seized domain name. A nervous website owner would probably prefer letting that website die out and starting a new one elsewhere. This induces evasion. It is therefore unsurprising that these websites have the feel of criminality; the law treated them as criminals. The forfeiture provisions of the PRO-IP Act undercut the cooperative spirit of the DMCA, when applied against websites.
All in all an excellent and worthwhile read for those interested in the legal issues here.