Credit Card Providers Aren’t Liable for Third Party Infringement–Perfect 10 v. Visa
By Eric Goldman
Perfect 10, Inc. v. Visa International Service Association, No. 05-15170 (9th Cir. July 3, 2007)
The Ninth Circuit has completed a hat trick of appeals involving Perfect 10′s litigation frenzy over online infringement of Perfect 10′s copyrighted photos. The resulting troika of opinions has important implications for cyberlaw and IP law, but the aggregate effect is hardly clear. The Ninth Circuit is visibly struggling to define and justify the boundaries of liability online, making it very hard to sort through the tangled mess of opinions.
Despite the problems with the specific details, the Ninth Circuit generally reached the right results in all three opinions. Financial service providers (FSPs) got a complete victory (12b6 dismissal), other support service providers (like ccBill) got significant protection, and Amazon/Google avoided direct copyright infringement but could be contributorily liable if they have sufficient involvement in the infringement (this liability is a little dicey, but we’ll see on remand). The opinions generally make a lot of sense if we focus only on the results.
In contrast, wading through the details shows just how problematic Ninth Circuit cyberlaw jurisprudence has become. The Ninth Circuit has chunked a few major Internet cases–Napster and Brookfield are two conspicuous examples–which has produced a long list of tortured subsequent precedent. The Ninth Circuit should bite the bullet and wipe those cases (and their progeny) off the books. Otherwise, the Ninth Circuit will torture itself each time it tries to reconcile new opinions with those erroneous opinions.
FSP Control Over Infringement
Online secondary liability cases generally struggle to define how much control over third party behavior is required to find secondary infringement. Here, the majority (Smith and Reinhardt) and dissent (Kozinski) philosophically disagree over FSPs’ ability to control infringing websites. FSPs have various rules prohibiting payment for illegal activities, and they do a lot of work to police their network. They could be made legally responsible for online gatekeeping (just like Congress did with respect to online gambling).
At the same time, the majority in this case sees FSPs as attenuated from the ultimate infringing acts. Like power companies, FSPs are behind-the-scenes vendors that don’t touch the flow of infringing bits. To the majority, that’s dispositive in this case. In contrast, Kozinski thinks FSPs may be no different than bagmen for an illegal deal, and they should take responsibility accordingly.
Like so many courts before it, this court cannot reconcile its competing characterizations–nor does it even hint at a doctrinal framework for developing a shared agreement on the matter. As a result, I think the law is still being decided case-to-case, substantially undercutting the precedential utility of this case.
The Legal Discussion
A few interesting aspects in the legal analysis:
* The court helpfully acknowledges the proliferation of contributory infringement tests and tries to harmonize them, saying prior articulations are “noncontradictory variations on the same basic test, i.e., that one contributorily infringes when he (1) has knowledge of another’s infringement and (2) either (a) materially contributes to or (b) induces that infringement.” I think this properly states the legal standard, including making it clear that inducement is a subset of contributory liability, not a separate test. Courts would be well-served to start with this language rather than the many other variants.
* The majority distinguishes the precedent by saying FSPs aren’t liable for contributory infringement because they “do not help locate and are not used to distribute the infringing images.” Kozinski’s dissent hammers the majority for its view, pointing out that payment systems are much more of a “sine qua non” for infringing websites than search engines. (He doesn’t use the Latin phrase, instead saying that “If cards don’t process payment, pirates don’t deliver booty”…which reminds me of another court’s judicial notice that booty is slang for buttocks). The majority’s response–websites could find alternative payment mechanisms–is true in theory but laughable based on current practices.
* The state claims, per ccBill, should be wiped out due to 47 USC 230, but 230 isn’t mentioned at all. Instead, the court rejects liability based on California common law precedent insulating FSPs from aider/abettor liability.
Where Are We? And Where Are We Going?
I wish I knew. We’re stuck in a legal limbo; as Kozinski concludes his testy dissent, “the opinion will prove to be no end of trouble.” I couldn’t agree more. The Ninth Circuit should take a very hard look at its entire body of cyberlaw if it wants to provide the type of useful guidance we expect from appellate courts.
On the other hand, I go back to the fact that the Ninth Circuit reached mostly sensible results in all three Perfect 10 cases, suggesting that maybe intuition is more useful at predicting the law than the court’s specific words. Relying on common sense isn’t very comforting, and the Ninth Circuit can and should do better, but perhaps common sense still plays a role in Ninth Circuit cyberlaw jurisprudence.