Case Western “Signifiers in Cyberspace” Conference Recap

By Eric Goldman

In mid-November, I attended a conference at Case Western Reserve University School of Law in Cleveland, Ohio entitled “Signifiers in Cyberspace: Domain Names & Online Trademarks.” My notes:

David Fewer spoke about Canada’s WHOIS policy. The old Canadian registry policy published registrant information without restriction. Then, the registry proposed a new policy not to publish personal information in the WHOIS database for individual registrants and for organizations that can show harm from publication. To reveal registrant information in those situations, a warrant would be required. That policy got amended to allow warrantless access for cybercrime enforcement, registered IP infringement and ID theft. Fewer argued that the amended policy violates Canadian privacy laws (PIPEDA) because consumers are not given adequate disclosures, the exclusions from the privacy policy are arbitrary, and consumers aren’t given the required option not to participate.

Corynne McSherry of EFF discussed how TM owners are bypassing direct challenges against gripers and instead putting pressure on domain name registrars. She focused on the Yes Man spoof website of the New York Times, which included a parody ad of the De Beers diamond manufacturer. Humorless De Beers sought relief from, the parodist’s registrar. EFF has responded to De Beers that the parody is legitimate because it has no commercial aspect, it’s nominative use, and the First Amendment applies. The EFF is also encouraging to ignore De Beers because it (as the registrar) can’t be liable for the registered domain name. So why is even entertaining De Beers’ complaint? Corynne notes the registrar’s revenue from any single domain name registration is less than legal cost of investigating and responding. Corynne discussed how parodists and gripers can minimize their legal risk (I blogged on these recommendations in May).

I remain very interested in situations where domain name registrars apply their own takedown policies to their customers. For example, I’ve previously mentioned GoDaddy’s “itchy trigger finger” when it comes to intervening with its registrants. I suspect there is significant heterogeneity among registrars’ interventionist tendencies. I think this is an area worth exploring. If you have other examples of domain name registrar intervention in its customers’ content, please share them.

Stacey Dogan spoke about the aftermath of the Rescuecom ruling. Stacey is disappointed that courts aren’t adopting her arguments to use the “trademark use in commerce” doctrine to insulate intermediaries (she calls it her “biggest failure in life”). She described three post-Rescuecom uncertainties: (1) what acts by intermediaries constitute TM infringement? (2) on what doctrinal basis? (direct v. contributory), and (3) what remedies do the intermediaries face?

Stacey thinks courts need to be more precise about the nexus between defendant behavior and TM owner harm. This should lead to better distinctions between direct and contributory infringement.

She offered a taxonomy of claims against intermediaries:

* General confusion = when the intermediary creates confusion through the blurring of ads and editorial content. Stacey thinks these aren’t TM issues. But if commingling is the problem, then the remedy should be an injunction requiring the intermediary to label the ads.

* Strict liability = when the search engine is automatically on the hook for its involvement with the ads. Stacey says courts should reject this approach due to the search engines’ lack of proximate causation for consumer confusion. If a search engine faces any liability, it should be solely on the basis of contributory infringement (with its higher scienter bar).

* Failure to act = when the search engine fails to respond to TM owner’s takedown notice. She said we don’t see this in search engine cases [a point I disagree with given that the TM owner vs. search engine lawsuits all represent a failing of the search engines’ voluntary TM policies]; instead, she was thinking of the Tiffany case. Stacey thinks the failure of act prong is where the legal action should be. She wants courts to map out appropriate scienter levels. General knowledge of infringement isn’t enough, and courts should let defendants make reasonable judgments about whether the advertiser will qualify for any trademark defenses. If the advertiser is obviously infringing, and intermediary gets notice and fails to act, she thinks contributory liability could be appropriate.

Graeme Dinwoodie believes the ECJ will not follow the Advocate General’s opinion in the Google case. He explored two parallels between the AG’s opinion and Rescuecom: Both get away from trademark use of commerce, and both consider underlying policy values. Graeme thinks search engine defendants should move away from disputing the lack of harm to the trademark owner; instead, he thinks they will get more traction by showing the countervailing benefits of their advertising. For example, he thinks they should be showing how keyword advertising can facilitate investment and innovation.

Jeffrey Samuels shared his perspectives as a panelist in 200 UDRP proceedings. Since the UDRP’s implementation, there have been about 25,000 UDRP decisions. 40% are US registrations. 75% involve .com. 75% are defaults.

The UDRP isn’t designed to solve all domain name disputes. He gave an example of a domain name registration containing a celebrity child’s name. The UDRP isn’t helpful because a 2 week old kid doesn’t have protectable trademark rights.

“The UDRP is hardly a model of clarity.” All cases are fact-dependent. If a UDRP proceeding has unusual facts, he recommends requesting a 3 member panel–these proceedings get more carefully evaluated opinions and minimize the effects of any one panelist’s idiosyncratic views.

Some issues that regularly arise in UDRP proceedings:

* What the TM owner has to do to establish its rights. The majority view is that a registration anywhere in the world suffices. Common law rights generally require presenting sufficient evidence validating the rights.

* There remains a split of authority on “sucks” sites.

* In the early days, panelists used to run through the multi-factor likelihood of confusion factors. That’s rarely done today. Now, most panelists just make sight and sound comparison.

Karl Auerbach discussed two interrelated issues: (1) ICANN lacks any political authority for its “Internet governance” role, and (2) technology does not require that ICANN monopolize DNS root services. He argues that we would benefit from competition among DNS root services. His argument reminds me a bit of the net neutrality debate. We can hypothesize many possible net neutrality problems, but most of them go away with vigorous competition. Similarly, ICANN’s often-ridiculous shenanigans would be less vexing in the face of bona fide competition for DNS root services.

Dan Hunter spoke about a new paper he’s writing with Mark McKenna. Their target is the fundamental trademark principle that trademark law protects against consumer confusion. They think consumer confusion is an imperfect proxy for our normative goal of protecting consumers. Some confusion is endemic in a complex society; and some methods of communication, like humor, require confusion to work. Therefore, they want to move away from trying to block consumer confusion and instead refocus trademark law on reducing errors in consumer decision-making. This seems like a fruitful endeavor, but they are also taking a swipe against the consumer search cost justification for trademark law, a move I didn’t follow.

Bill McGeveran recapped his recent work on social networking sites and gave a preview of his next article. His target is fake online profiles such as the Tony La Russa fake Twitter account. He expects to see more pressure to create IP rights in personal identities.

I spoke about trademarks and behavioral targeting, and in particular the competition among marketers for consumer preference information. For example, I believe the anti-deep packet inspection pushback wasn’t based solely on privacy concerns. Instead, destination websites fear that an IAP will disintermediate them and use its prime access to consumer preference information to steer customers to competitors. (See this blog post for more on that point). My (very brief) slides.