Tiffany, blue

US IP blogs are abuzz with the Tiffany’s v. eBay case. It’s certainly a major victory for eBay, and by extension perhaps numerous other internet services. The nominative fair use determinations for eBay’s on-site banners and off-site search engine optimization are particularly welcome. The decision on secondary liability seems well-reasoned to me. Eric Goldman does a general reaction post that touches all the bases.

I do want to point out, however, just how much Tiffany’s got out of eBay by pressure, in particular eBay’s willingness to prohibit “Tiffany style” and “inspired by Tiffany” in ads (with, as best I can tell by searching on eBay’s site, an exception for a “Tiffany-style setting” for diamond rings). This is troubling because “Tiffany style” is in itself an acceptable nominative use for certain products that are like, but are not, Tiffany jewelry. The presence of the word “style” means that initial interest confusion, that despicable animal, is unlikely. And the possible claim that “Tiffany style,” while not confusing to the buyer, would lead to post-sale confusion depends on the argument that the design of every piece of the underlying jewelry is protected by trade dress, an argument Tiffany’s did not make in this case. Of course, it may well be the case that “style” was so often a signal of counterfeit status that a ban was in order, but the evidence recited by the district court didn’t go to that.

The lesson is that eBay, like other internet aggregators, is interested in its own welfare, not in the maximum efficiency of the competitive system or in justice. So it will fight overreaching trademark claims precisely to the extent that it makes business sense to do so, and no further, just as our beloved Google prohibits even nominative fair use of other parties’ trademarks in the text of sponsored ads. (Speaking of our beloved Google, Scott Westerfeld’s recent novel So Yesterday has a narrator who’s sworn off mentioning famous brands in order to fight their famousness, but even he makes an exception for Google, because, he says, his story wouldn’t get very far without the ability to use the verb—and the search engine. That’s brand power, and it’s also why genericide has functionally disappeared for fanciful marks.)

Now, a bit about dilution: (1) The court adopts the interpretation that dilution requires that a defendant to use the mark to identify its own goods or services, which eBay didn’t do. (2) I found it amusing that the court’s careful discussion of the potential divergence between state and federal dilution law created by the very different wording of the TDRA ends with the same time-saving conclusion as every other court: might as well consider them together! And really, who’d want to do two dilution tests? (Who’d want to do one?)

Last, because my niche is advertising law, a word about the false advertising claims: The argument was that eBay advertised the availability of Tiffany jewelry, but, given the high percentage of counterfeits for sale on the site, that was as false as advertising that a model of car gets 35 mpg when only 25% of the actual cars do so. The court rejected this argument, and the basic intuition is similar to the one that’s hampered post-Dastar claims about false attribution as false advertising: when one’s core claim sounds in trademark, §43(a)(1)(B) should not serve as an end run around the requirements of trademark law.

Since authentic Tiffany jewelry was for sale on eBay, the claims were not literally false even if a buyer was more likely than not to get a counterfeit piece. Comment: Were it not for the trademark overlap, I’d find this a dubious conclusion, given the car analogy above. But, as the court pointed out, it had already found that the use of “Tiffany” was a protected nominative use.

In terms of implicit falsity, the court held that Tiffany’s claim depended on eBay’s knowledge of high counterfeiting levels. Again, as false advertising doctrine, this is wrong—false advertising is strict liability. But it is common in §43(a)(1)(B) cases brought over threats of patent infringement to hold that only knowingly false claims made about patent rights can be false advertising, while sincere but wrong claims are not actionable. Intent/knowledge here play the role of boundary markers more than elements of a false advertising cause of action: these subjective elements manage the scope of patent rights, or trademark rights, and are not true measures of whether false advertising has occurred in the abstract.

Finally, the court found that the false advertising was sellers’ responsibility, not eBay’s. Again, I don’t think much of this, to the extent that eBay itself chose to advertise the term “Tiffany & Co.” However, insofar as the court was suggesting a secondary liability analysis for false advertising in the context of online auctions, that makes more sense—just as eBay is not directly liable for trademark infringement by counterfeiters, it’s not directly liable for false advertising by sellers of bad merchandise.

The conclusion “it cannot be said that eBay was misleading customers when eBay was diligently removing listings from the website that were purportedly counterfeit” sounds factual, but it’s actually normative: it is logically possible and even plausible that buyers were misled despite eBay’s policing (and perhaps in part because of it)—indeed, the record contained numerous emails from buyers complaining about counterfeit Tiffany sales that slipped through the screening; such buyers were materially misled by the listings’ claim to offer authentic Tiffany jewelry. Nonetheless, the court was unwilling to impose liability. One way to think about this might be materiality: given the availability of authentic Tiffany merchandise on the site, there wasn’t a sufficient connection between eBay’s general advertising that Tiffany items could be found on the site and any particular auction listing.