Ad Networks Ordered to Drop Allegedly Infringing Site–Elsevier v. eNom
By Eric Goldman
On the surface, this seems like a run-of-the-mill copyright enforcement. The plaintiffs Elsevier and John Wiley publish textbooks on pharmaceuticals and related topics. The website at issue, Pharmatext.org, allegedly republishes those copyrighted textbooks for free via an ad-supported website. If these were the only key facts, the publishers should be able to shut down the rogue site easily.
However, I think this is a pretty complicated case that isn’t getting the nuanced legal analysis it requires. There are at least three major complexities to the case:
1) Pharmatext.org is now offline, but when I reviewed it in Google’s cache, it looked like a linking site. In other words, the site itself wasn’t hosting the allegedly infringing downloads but just linking to URLs where they were available. I didn’t see anything in the complaint that connected the operators of Pharmatext to the uploads of the copyrighted material on third party servers.
Thus, this case could be about a website’s liability for linking to infringing content. We might still conclude that Pharmatext is infringing, but we’d need to reach that through a secondary liability analysis, not a direct liability analysis. Of course none of this is mentioned in the materials I saw. The complaint does not clarify who is the direct infringer because those paragraphs are in the passive voice–is it the uploader, the individuals who download, or Pharmatext for unspecified activities that it did?
2) Pharmatext.com is owned by eNom as a privacy proxy (under its Whois Privacy Protection Service, the captioned defendant). I’ve previously blogged about domain name privacy proxies before and the legal troubles they are encountering (see my post on Solid Host v. NameCheap). The publishers sue WPPS for vicarious copyright infringement, alleging “it controls the domain name pharmatext.org and receives a direct financial benefit for doing so.”
Hold on a sec. This is a really dense statement that requires unpacking. WPPS is the domain name registrant on behalf of an undisclosed principal. It does not “control” the domain name. Even if it did, the normal test for vicarious infringement requires the “right and ability to supervise the infringing activities.” How does a privacy proxy do that? It controls the domain name, not the associated servers. See the old 9th Circuit ruling in Lockheed v. NSI. And typically vicarious infringement requires a direct financial benefit from the infringement. Sure, as a service provider, the privacy proxy gets paid by the website operator–but the payment amounts don’t vary with the amount of infringement. This is like saying the electric company gets paid by an alleged infringer to supply power to the infringing website. Yes, the electric company gets paid, but no, that’s not direct financial benefit from the infringement. (Ironically, in fact increased infringing activity might boost the amount of electrical consumption, so the electric company is more likely to profit from infringement than a privacy proxy.)
Finally, recall that I believe Pharmatext is a linking site that isn’t the direct infringer. So WPPS is a distant service provider to a possible secondary infringer. Tertiary liability, anyone?
The court’s response? It orders WPPS to disable the domain name, fork over the principal’s identity, and freeze any transfer of the domain name.
I am extremely bearish on the future of domain name privacy proxy services. It seems inevitable that IP plaintiffs are going to drive that particular service offering into the ground with their litigation.
3) The publishers also sued two ad networks, Chitika and Clicksor, for contributory copyright infringement. The supporting allegations? The ad networks directly profit from the infringement and provide the funds to enable Pharmatext.com to continue its existence.
Traditionally, contributory copyright infringement requires knowledge of the infringing activity and a substantial contribution to the infringing activity. I didn’t see any allegations of knowledge by the ad networks at all–no assertions that the ad networks had figured out that Pharmatext,org was a rogue website, and no assertions that the copyright owners sent a C&D or takedown notice. (I don’t think the ad networks qualify for a 512 safe harbor because of the way those safe harbors are worded; but a C&D/takedown notice would still help the publishers in arguing that the ad networks knew it was a rogue website). The ad networks’ only “contribution” to the infringement is the payment of money to the site, and this was expressly rejected as sufficient for liability in Perfect 10 v. Visa, which concluded that even if the payment system stopped the flow of money, it would not automatically stop the website and any associated infringement on it.
Further, if Pharmatext was a contributory infringer for linking to infringement, this means the ad networks would be contributing to a contributory infringer–another tertiary liability argument.
The court’s response: it ordered the ad networks not to pay any money to Pharmatext and to drop them as customers.
From the PACER report, it appears that the defendants haven’t contested the publishers’ claims. I hope that will change. This is a case where the publishers have clearly overreached, and if the judge won’t call them on it himself, we need the adversarial system to expose the major gaps in the publishers’ logic.
Meanwhile, this case illustrates two broad themes. First, it illustrates how plaintiffs are going after an array of supporting service providers to make them responsible for their customers’ activities. On the trademark front in the past 12 months, see, e.g., Louis Vuitton v. Akanoc (web hosts), Gucci v. Frontline (payment service providers), Roger Cleveland v. Price (web designers/SEO), Microsoft v. Shah (software vendors that assist website development). Plaintiffs are reading secondary liability doctrines very, very broadly, and courts aren’t always slamming down those arguments as emphatically as they should.
Second, several other lawsuits have tried to nail ad networks for providing advertising support to infringing sites. As another example (also in the trademark realm), see the Vulcan Golf v. Google lawsuit. The only successful case that I can recall involves Triton Media, a ruling that was so skewed by its facts that I didn’t think it was worth a full blog post. Wendy Davis discusses the comparison more.
If this lawsuit ends with the current injunction, I’ll consider this ruling an interesting oddity. Any further developments in this case warrant careful scrutiny.