Video Embedding Site Isn’t a Contributory Copyright Infringer, But Sideloading Could Be Direct Infringement–Flava Works v. myVidster
By Eric Goldman
myVidster is a “social bookmarking” website that allows users to link to videos hosted elsewhere on the Internet and thereby embed the videos in myVidster’s user interface. Today, myVidster scored a big win at the Seventh Circuit, which held that it had not committed contributory infringement by allowing users to embed infringing videos via myVidster. It’s hard to state just how amazing this ruling was for myVidster, because myVidster’s principal, Gunter, often refused to honor takedown notices (on the dicey premise that anything posted somewhere elsewhere on the Internet was freely linkable) and thus presumptively failed to qualify for the 17 USC 512(d) safe harbor. Normally, when a website fails to honor takedown notices, judges come down hard on the website—just like the district court did in this case.
While the opinion offers good news for myVidster and possibly other linking websites, it does raise a concern about sideloading, i.e., grabbing a remote file that a user links to and making an archive copy of that file for further delivery. The opinion, without much elaboration and without using the term, breezily characterizes sideloading as direct infringement by the website for grabbing and republishing the remote file. This is almost certainly bad news for Pinterest, which (I believe) routinely engages in sideloading without a lot of explanation to its users, presumably premised on the idea that storing the remotely linked files is authorized by the user and thus qualifies for the 512(c) safe harbor. It remains to be seen what will happen if Pinterest’s sideloading is directly challenged and Pinterest is around to defend its interests, but this ruling provides a warning that judges may not see it the way Pinterest does.
Judge Posner drafted this opinion, which means that (as usual for him) it reads like a barely edited first-draft. As usual for Seventh Circuit opinions, it makes a number of questionable and undefended offline analogies, makes assumptions about factual questions that could/should be remanded to the district court, barely engages with or cites to other legal precedent, raises and addresses issues that the litigants never raised, and is filled with gratuitous digressions (e.g., an uncomfortable discussion that gay ethnic pornography might be illegal, a contention neither party advanced; and an odd discussion about the reputational capital benefits of sharing content). If I were Flava Works, I would be hopping mad about the manifest procedural defects in the opinion (and motivated to seek en banc review). As a result of the opinion’s characteristic affectations, parsing this opinion is needlessly difficult, so I can only do so much to deconstruct the legal principles in it.
The main ruling is that the folks uploading infringing videos are direct infringers, but myVidster isn’t contributorily liable for letting users link to those infringing videos. Posner unhelpfully rejects the standard Gershwin definition of contributory infringement; instead, he idiosyncratically defines it as “personal conduct that encourages or assists the infringement.” (In my opinion, the last thing we need is further proliferation of definitional standards for secondary infringement!). myVidster wasn’t responsible for any infringing activity mostly because it didn’t copy or distribute the infringing videos or help users copy or distribute the infringing videos.
Posner sidesteps the concerns from amici Google and Facebook about possible “tertiary” liability, saying myVidster didn’t commit contributory liability and that ends the inquiry. And because the opinion doesn’t find any secondary infringement, it says the 512(d) safe harbor for linking to infringing content is not needed. That turns out to be good news for myVidster because it had a low chance of succeeding with a 512 defense. myVidster avoided an inducement claim because it didn’t “invite” its users to link to infringing files.
Posner does separately address the 106 public performance right, but I found the opinion about that especially inscrutable. I’m hoping someone else can help me understand what Posner was trying to say and, in particular, why he discussed Fonovisa (an offline swap meet case) when discussing the performance right. Although the opinion cited the Ninth Circuit’s Perfect 10 v. Amazon ruling, I also don’t understand why Posner did not discuss the Ninth Circuit’s thoughtful and highly relevant discussion about public display and the difference between embedding content via links and hosting it. The only thing that is clear to me is that Posner thought that linking to infringing content is just like citing to someone else’s content, and thus myVidster was too remote from any infringing activity to be responsible for it.
[Rebecca provided this helpful distillation of Posner's public performance discussion: "the site hosting the unlawfully copied videos is publicly performing them, but the watcher is not, and therefore assisting the watcher (without hosting the video) does not assist copyright infringement." If that's where he was going, it would have been interesting to see Posner analogize to and distinguish the Cablevision case.]
The district court’s ruling against myVidster reflected, in part, Posner’s own opinion in Aimster, which basically threw a P2P file-sharing service under the bus. I personally think Aimster was one of numerous P2P exceptionalist cases and therefore only has precedential value in the P2P file-sharing context. Here, Posner distinguishes his own Aimster opinion in a way that strongly supports that interpretation, even though he doesn’t explicitly acknowledge that implication.
In particular, he invokes the troublesome methodology of determining a service’s liability based on the percentage of infringing activity taking place on it, saying most (all?) files on Aimster were infringing but that wasn’t the case for myVidster. Comparing infringement rates is a realpolitik approach, but one fraught with peril because (a) those ratios can and do change over time, and (b) they are notoriously difficult to measure properly. Worse, here it appears Posner (or possibly his clerks) did his own ad hoc empirical assessment of myVidster’s infringement rates–on July 4, no less. Your honor, like the American public you serve, you deserve to enjoy the holiday! If the outcome will turn on the current percentage of infringing activity—which it shouldn’t—that question should be remanded to the district court for further fact-finding, not by having a septuagenarian surfing for gay erotica (on court computers?) on a national holiday.
Posner also doesn’t renounce his ill-considered discussion Aimster about the service’s name, where he castigated it for using the –ster suffix just like Napster. Of course, myVidster also uses the –ster suffix, but to no legal consequence this time. Another reason to relegate Aimster to the P2P exceptionalist branch.
While the secondary infringement ruling is favorable to myVidster, the opinion indicates that myVidster could be enjoined from allowing users to archive the linked videos as part of a paid membership service. myVidster had already discontinued that premium service, so the opinion doesn’t spend much time discussing it. Nevertheless, the opinion flatly says that the archiving videos from other sites qualifies as direct infringement, not secondary infringement, even if it’s presumptively initiated by the users themselves as one of the benefits of a premium membership. I think it’s erroneous to treat sideloading as direct infringement when it’s done at a user’s direction, but chalk that up as another thing this opinion got wrong.
This opinion is not an instant classic. This opinion touches on so many important questions in secondary copyright jurisprudence—e.g., the 512(d) safe harbor, liability for tertiary infringement, the legitimacy of UGC linking sites that often link to infringing files stored somewhere else, sideloading and more. Unfortunately, because the opinion’s drafting is a train wreck, I’m skeptical it will be all that important or influential. As a result, it’s a big win for myVidster and less helpful for everyone else.
Nomenclature watch: I believe this is the first appellate opinion to use the term “pay wall” (or “paywall”). The only other opinion I found using either term in Westlaw or Lexis was the district court opinion in this case.
UPDATE: Mike Masnick sees more good news in the opinion than I do.