Ad Network Avoids Contributory Copyright Infringement for Serving Ads to a Rogue Website–Elsevier v. Chitika
By Eric Goldman with comments from Venkat
As Venkat recently indicated, while Congress generates massive amounts of hot air discussing SOPA, there’s a common law battle quietly brewing with rightsowners seeking to obtain in court the same basic remedies that SOPA would provide statutorily. This judge nixes this rightsowner’s request, but this is hardly a win for SOPA opponents or ad networks.
Elsevier publishes textbooks. This lawsuit involves pharmatext.org, a (now-defunct) site that provided links to allegedly infringing copies of Elsevier’s books. We’d now characterize pharmatext.org as a possibly rogue website. I previously blogged about this case in January, when the judge issued a dense and troubling ruling that ordered ad networks Chitika and Clicksor to freeze the site’s money and stop serving ads to the site. It also ordered the domain name privacy proxy to freeze the domain name and identify the owner. Basically, Elsevier got much of the relief that SOPA now seeks to enact as law.
But hold on a second. The court’s January order was based on ex parte proceedings. Chitika subsequently showed up to contest the case, and surprise! The court reaches a different result after adversarial proceedings. Let’s hear it for due process!!! YEAH!
[Note: you may recall Chitika separately had an unfortunate legal scrape with the FTC.]
Chitika first argued that Elsevier never proved any direct copyright infringement in the United States because Pharmatext.org’s operator, Saggi, was based in India, and the site had no US presence. Elsevier argued that it had test downloads done by US investigators, so there was infringing activity in the US. The court nearly blows a gasket before punting this sticky jurisdictional issue:
While it appears that Chitika may eventually be entitled to judgment on this ground (that is, plaintiffs’ failure to allege any act of direct infringement occurring entirely within the United States), factual issues involving the structure of the Internet and the locus of the infringing activity remain (Where did the copying take place? Where are the third-party websites and servers, from which unauthorized copies of plaintiffs’ books were downloaded?). These issues preclude the granting of the motion on this ground.
For more on the jurisdictional question, see, e.g., the Shropshire case. This locus-of-infringement question is quite interesting, especially as applied to “foreign” rogue websites, but SOPA largely sidesteps all of the doctrinal complexity (and creates its own mess).
Without resolving whose the underlying direct infringer, the court instead considers Chitika’s contributory copyright liability. The court concludes that Chitika doesn’t have the requisite knowledge of infringement:
Plaintiffs do not allege facts showing that Chitika was familiar with the content of the Pharmatext website, or knew (or had reason to know) that such content was infringing. Thus, plaintiffs fail to support with plausible facts their conclusory allegations that Chitika “must have had knowledge” of the alleged infringement of plaintiffs’ books…and that Chitika “plac[ed] ads on the Pharmatext site because [it] believe[d] that Pharmatext users – in other words, people seeking to obtain pirated copies of copyrighted books – are a target audience for particular advertisers.”
Notice this leaves open what would have happened if Elsevier had sent a takedown/cutoff notice to Chitika. Presumably, such a notice would have conferred the requisite knowledge to Chitika.
Heavily relying on Perfect 10 v. Visa (even though it wasn’t binding precedent in Massachusetts), the court also rejects Chitika’s material contribution to the infringement. The court says:
while Chitika’s advertising payments might make it easier for Saggi’s infringement to be profitable, Chitika did not create, operate, advertise, or promote the infringing websites, and its advertisements were not the “site” of the infringement.
However, the court hedges at the end, saying it wasn’t deciding the issue definitively because Chitika’s lack of knowledge resolves this case.
In a footnote, the court explains why Judge Kozinski’s dissent in Perfect 10 v. Visa is unhelpful to the plaintiffs:
in contrast to Visa, Chitika did not provide an “essential” service to Saggi that enabled infringement on a “massive scale.” Plaintiffs make no factual allegations that Chitika knew about any infringing activity, nor is there any evidence that Chitika was “intimately and causally involved in a vast number of infringing transactions.”
This is a little garbled (because it references the defendant’s knowledge in the consideration of material contribution), but it seems like the court suggests that ad networks may be in a different position than payment service providers because ad networks don’t have the same multi-iteration “transactions” with the rogue website.
Notice that this court totally sidestepped (or missed?) the tertiary liability aspect of this case–that Chitika was a support provider to a site that only provided links to allegedly infringing files. To me, it would be entirely appropriate for the court to say that any tertiary player categorically lacks the ability to materially contribute to infringing activity. Otherwise, once we start doing a dragnet for service providers to service providers to infringers, the universe of potential defendants grows to a ridiculous size.
In the end, the court grants Chitika’s Rule 12(c) motion for judgment on the pleadings. Broadly construed, this ruling is a win for ad networks, indicating that they are not automatically liable for contributory copyright infringement simply because allegedly rogue websites participate in their networks. But I’d hardly call this a resounding win. Elsevier didn’t send Chitika a cutoff notice, giving the court an easy escape valve on scienter, and the court waffled on the material contribution prong. I would expect this opinion to look very different if Elsevier sent a cutoff notice and Chitika didn’t promptly drop Pharmatext. In so, rightsowners like Elsevier probably can get 90%+ of the benefit of SOPA Section 103 simply by sending cutoff notices to ad networks.
Notice also that in the face of a cutoff notice, Chitika would not stand up to defend the alleged rogue website publisher in its network. Chitika alleged that Saggi accrued about $500 in royalty payments over 29 months of service. Assuming Chitika does 50/50 splits with its publishers, Chitika will not expend an ounce of effort to preserve its $17/month revenue stream from Saggi. Thus, Elsevier’s cutoff notice would be dispositive–even if Chitika could win a ruling like this (which would be more uncertain after Chitika gets a cutoff notice), it’s not worth the fight. So after Elsevier’s cutoff notice to Chitika, Chitika instantly tosses Pharmatext overboard like a piece of garbage, due process be damned. SOPA isn’t required to get that result.
Comments from Venkat:
As Eric points out, Elsevier is trying to hold Chitika liable under a theory of tertiary liability. It’s far from clear that Pharmatext is liable to Elsevier for merely linking to infringing downloads. Columbia Records v. Fung is currently pending in the Ninth Circuit, and it should shed some light on how extreme the facts need to be in order to hold a website liable for merely linking to allegedly infringing downloads. (Here’s Eric’s post on the district court opinion in that case: “Torrent Sites Induce Infringement and Lose DMCA Safe Harbor–Columbia v. Fung.” I expect the Ninth Circuit’s ruling on the appeal fairly soon.)
The question of infringement that occurs purely off-shore is interesting. As Eric mentions, the court questions whether a site that is totally off-shore can be held liable if there’s no predicate act that occurs in the United States. I wonder if this is a legitimate argument for rightsowners as to something in the rules that should be ‘fixed’? If Elsevier has no recourse in US courts against Pharmatext, should this automatically undermine any possibility of going after service providers of rogue sites?
Finally, as to the issue of whether Chitika has the requisite knowledge of the underlying infringements for Elsevier to be able to hold it liable, query as to whether any search terms that were used to target could be used to show the requisite knowledge on Chitika’s part. The use of search terms as a proxy for intent has come up in the copyright context (e.g., it was used against Limewire as indicative of its intent to lure Napster users) but it would seem less probative in this context.