BitTorrent has been in the (copyright) news lately – and not surprisingly – after the movie studios set their sites on bringing down yet the latest iteration of file-sharing technology.
2 great background sources on what BitTorrent is and how it works can be found here and here. In short terms, BitTorrent is a file sharing technology, different from Napster and its peer-to-peer progeny in that it draws down pieces of large data files from multiple computers – rather than single computer to single computer peer-to-peer – based on a “community” structure of participating individual users. The two biggest distinctions are (1) no single source for the compiled total file contributes more than a very small portion of the total file and (2) the distributive structure finesses the constant file-sharing problem of large data transfers demanding large broadband resources.
Why is bitTorrent in the (copyright) news?
BitTorrent is in the news not simply because Netflix’ CEO stated that “we’ve finally beaten bitTorrent.” (“We”, by the way, presumably refers to Netflix’ full-file streaming capabilities.)
BitTorrent is mostly in the news because of its enormous popularity among file sharers, which of course begs the question why has it garnered so much popularity among file sharers? Among other reasons, Napster and Grokster are gone, Limewire just recently settled all sorts of lawsuits with music labels, and the next target of the movie studios and content owners is … bitTorrent.
Legal issues aside (at least until the next paragraph), bitTorrent is an innovative exploitation of the internet’s networking and cloud-computing capabilities to distribute large and ever-larger files seamlessly and quickly. From a the standpoint of a non-technologist observer, it is simply pretty neat how it takes advantage of distributive concepts of sharing and cost- and resource-burden shifting.
From the standpoint of content creators and intellectual property owners, however, the technology is really quite law-evading.
The isoHunt case
A leading bitTorrent case is in federal court in California, involving Columbia Pictures and other film studios who sued a bitTorrent company called isoHunt (defendants were isoHunt and its founder, Gary Fung). The defendant – isoHunt – hosts a bitTorrent service, which was challenged by several major film studios as serving no purpose other than facilitating direct infringement by the technology’s users of copyrighted films and television programs. And according to the film studios, doing so intentionally and willfully.
Columbia Pictures moved for summary judgment, and the court granted it, finding that isoHunt was liable to the studios for secondary copyright infringement. To be clear, Columbia argued that the bitTorrent company itself didn’t directly infringe the copyrights, but rather improperly “induced” copyright infringement by its users. Columbia’s summary judgment motion was granted, and thus isoHunt was liable for copyright infringement.
The court’s summary judgment Order can be found here.
isoHunt appealed the case to the 9th Circuit Court of Appeals in San Francisco, and the appeal is pending. isoHunt’s appeal brief can be found here.
Relevant legal issues
Several important copyright arguments were made in the isoHunt case. Of most significance were 2 particular issues: Inducement of copyright infringement and the safe harbor for providers of “information location tools” under Section 512 of the Digital Millennium Copyright Act (the DMCA).
First, Inducement of Copyright Infringement
The film studios claimed – successfully – that isoHunt should be held liable for copyright infringement under a “secondary liability” theory.
This argument relied on the Supreme Court’s Grokster case, and its standard for determining inducement liability: “inducement requires that the defendant has undertaken purposeful acts aimed at assisting and encouraging others to infringe copyright”.
Grokster was the 2005 US Supreme Court case in which a peer-to-peer file-sharing service was held liable for copyright infringement under a theory of “inducing” infringement by others. In the isoHunt case, the court ruled that the plaintiffs – the movie studios – had proven inducement by the defendant’s bit-torrent technology, applying the Grokster test for inducement.
What facts supported this argument? Lots of them. Some examples (see commentary from Michael Barclay here):
- Expert witnesses for the studios used statistical samplings of the content and server logs to show that 90-95% of all available content available through the defendant’s service infringed copyrights.
- The studios offered evidence that isoHunt’s search code was expressly designed to find copyrighted material: As ArsTechnica reported, “One of [isoHunt’s] sites also displayed a list of the top-20 grossing movies in the US, with links to copies of each, while another had categories that included “High Quality DVD Rips” and “TV Show Releases.””
- The defendant’s own blog made statements questioning whether copyright infringement was really theft.
Second, DMCA safe harbor
The court then rejected isoHunt’s DMCA arguments based on the same facts, ruling that the same actions that demonstrated isoHunt’s inducement conduct also demonstrated either actual or presumed “knowledge” on the defendant’s part to nix any DMCA safe harbor argument.
For example, isoHunt argued that much if not most of the service’s offending content resulted from user queries that generated automatic responses – essentially, it was the user downloads of the dot-torrents that generated the infringing copying (or infringing distribution) of the copyrighted films and television shows. If that argument had been accepted, this would have safely placed isoHunt within the DMCA’s safe harbor for information locator services, shielding the service provider from liability for content posted by its users.
This was the really interesting part of the case, because this is where – at least from the position of more established companies like Google – the court went off the rails with its decision.
First: In a footnote, the court made this statement about dot-torrent file downloads:
It may be true that the act of downloading a dot-torrent file is not itself a copyright-infringing action; but once that dot-torrent file triggers the process of downloading a content file, copyright infringement has taken place. Because dot-torrent files automatically trigger this content-downloading process, it is clear that dot-torrent files and content files are, for all practical purposes, synonymous. To conclude otherwise would be to elevate form over substance.
isoHunt had argued that whether or not it may have done things to “encourage” or “suggest” copyright infringement, its core, basic actions as embodied in its technology – dot-torrent files – did not (because it could not) constitute copyright infringement. Thus, isoHunt made a sort of lack of “nexus” argument: Before you can claim inducement of copyright infringement, you have to prove … copyright infringement. And therefore, users’ actions in downloading the dot-torrent files were not copyright infringement at all, and therefore isoHunt shouldn’t be held responsible for having induced anything illegal.
As just quoted, the court made fast work of this argument.
Second, the court used this same point to illustrate how isoHunt had undercut its own DMCA safe harbor argument: “upon obtaining such knowledge or awareness [of infringing material], [the service provider] acts expeditiously to remove, or disable access to, the material.” (DMCA Section 512(d)). isoHunt had not complied.
Third, quoting from TechDirt: “Most safe harbor cases have focused on whether or not the service provider responded to notices, but in this case, the court said that there were enough “red flags” that, even in the absence of notices, IsoHunt should have blocked certain files.”
Why is this case in the news?
Short answer? Because Google got involved. Longer answer? Because this case was an unusual instance where a court ruled that DMCA safe harbor protection was not available to a provider of “information location tools” who knew or should have known about potential or actual copyright infringement happening on its service. On the appeal, Google – which was not a party to the case – sought to intervene in the case arguing that the lower court got it right when it held against isoHunt for inducing copyright infringement, but also that the court went too far when it further held that this same reasoning nixed the availability of the DMCA safe harbor. Or in other words, that inducement of copyright infringement in and of itself is distinct – and sufficiently culpable – as grounds for liability for copyright infringement.
Google’s brief on the appeal can be found here.
The court in isoHunt stated:
inducement liability and the Digital Millennium Copyright Act safe harbors are inherently contradictory. Inducement liability is based on active bad faith conduct aimed at promoting infringement; the statutory safe harbors are based on passive good faith conduct aimed at operating a legitimate internet business. Here, as discussed supra, Defendants are liable for inducement. There is no safe harbor for such conduct.
It is this statement that drew Google’s attention and the attention of many commentators (see, for example, these stories from arstechnica and paidcontent). Google is right in arguing that “mere knowledge of infringing potential or of actual infringing uses would not be enough here to subject a distributor to liability” (quoting from Grokster), and that a long line of cases had held that this “mere knowledge” hasn’t been sufficient to nix the availability of the DMCA safe harbor.
The problem for Google’s argument, though, is that that is what the DMCA safe harbor does actually state. So, and more to the point, the language that Google relies upon is from Grokster, which was a case involving claims of inducement of copyright infringement, not the availability of the DMCA safe harbor. Relying on language from Grokster – an inducement of infringement case – to address a DMCA safe harbor case may be misplaced.
Google argues that isoHunt’s lack of compliance with the DMCA’s takedown requirements is sufficient to nix any DMCA safe harbor defense that might be available. So that, any further discussion of whether the DMCA would otherwise be available to a service provider in this situation is inappropriate.
A conceptual problem or an actual problem?
A problem in bitTorrent cases is the lack of any potential non-infringing uses, making the perverse situation where the yet – as Google and other DMCA safe harbor advocates would have it – further evidence is still needed of the service provider’s affirmative, demonstrable steps acknowledging infringement. In isoHunt, arguably this wasn’t a big deal because you had a really bad actor doing egregious things.
On its appeal, isoHunt argues that it does the same thing that Google does when it comes to torrents, and that it is entirely easy to search for torrents on Google ((using a filetype: command for “torrent”). Google probably cannot argue otherwise, but it can argue against (a) the inducement activities that isoHunt does (see above) and (b) isoHunt’s nonresponsiveness to DMCA notices.
Add Comment