eBay Mostly Beats Tiffany in the Second Circuit, but False Advertising Claims Remanded

By Eric Goldman

Tiffany (NJ) Inc. v. eBay Inc., 2010 WL 1236315 (2d Cir. April 1, 2010)

In a subtle opinion with potentially significant implications, eBay has preserved most of its big 2008 district court victory in the long-running Tiffany v. eBay case. However, as seems to be the norm with federal appellate opinions, the opinion intentionally sidesteps some key open doctrinal questions squarely raised by the case–such as if the Second Circuit recognizes the nominative use defense, or the Second Circuit’s standards for contributory trademark infringement. As a result, we don’t get the clean and decisive doctrinal standards that help make a case truly precedent-setting; if anything, some aspects of the case are pretty eBay-specific. And while the opinion is generally a win for defendants, there are at least two tangents (the willful blindness standards and the false advertising discussion) that future plaintiffs will unquestionably explore.

What is clear from the opinion is that Tiffany has lost the battle to force eBay to change its system to do more for Tiffany. (Unless Tiffany successfully appeals to the US Supreme Court, which it has said it is considering). What isn’t clear to me is how much money is still on the table for the false advertising claim. If Tiffany can make this a backdoor way to get its alleged damages from counterfeits on the eBay site, then there’s still potentially a lot of money left to fight over. But otherwise, it would make way more sense for the parties to settle up rather than pursue the remaining false advertising issue.

Overview of the Rulings

* eBay isn’t directly liable for trademark infringement based on its advertising activities due to the nominative use doctrine (even though the panel did not adopt the doctrine)

* eBay isn’t secondarily liable for counterfeit sales because generalized knowledge is not enough and eBay followed a notice-and-takedown procedure

* eBay isn’t liable for direct trademark dilution based on the unadopted nominative use doctrine

* the appellate court rejected the lower court’s reliance on a nominative use defense to the false advertising claim.

The ultimate disposition: “The case is therefore remanded…for further proceedings for the limited purpose of the district court’s re-examination of the false advertising claim in accordance with this opinion.”

Direct Trademark Infringement

The Second Circuit has never expressly adopted the nominative use defense, one of the many reasons why I favor doctrinal gatekeepers like “use in commerce” rather than relying on the inconsistently interpreted defenses. This panel doesn’t adopt a nominative use defense for the Second Circuit either, but it does say “a defendant may lawfully use a plaintiff’s trademark where doing so is necessary to describe the plaintiff’s product and does not imply a false affiliation or endorsement by the plaintiff of the defendant.” This is pretty similar to the Ninth Circuit’s standard, though it omits the factor that the defendant took only as much of the trademark as was needed (in practice, that might be subsumed in the “necessary” requirement).

eBay qualified for the nominative use defense for both its website references to Tiffany and its use of Tiffany in sponsored links because “eBay used the mark to describe accurately the genuine Tiffany goods offered for sale on its website. And none of eBay’s uses of the mark suggested that Tiffany affiliated itself with eBay or endorsed the sale of its products through eBay’s website.”

Other than the district court opinion in this case, I can’t think of another case that has found a nominative use defense in a keyword advertising case. This ruling suggests that anyone who might qualify for a nominative use–certainly retailers and affiliates, but potentially even competitors–might have some additional traction to their defense. Potentially, this case will take some wind out of the sails of Rosetta Stone, which has partially complained about advertisements by folks in its channel. All of those people should be protected by the nominative use doctrine, and the Second Circuit may have just filleted those folks out of Rosetta Stone’s case.

Note that at least one of eBay’s ads was triggered by “tiffany” and included “tiffany” in the ad copy, so it appears that the nominative use defense extends to the trigger. The ruling does not address what happens if the trademark acts as a trigger but does not show in the ad copy. Before the 2nd Circuit’s April 2009 Rescuecom v. Google ruling, the trial judge held that triggering by itself didn’t qualify as a use in commerce. The appellate court doesn’t cite Rescuecom or address use in commerce at all; but I believe this is the first Second Circuit keyword advertising case since Rescuecom, and it turns out fine for the defense.

The court rejected Tiffany’s argument that eBay’s general knowledge of counterfeiting activity was relevant to the direct infringement inquiry. The court also sensitive to Tiffany’s apparent motive of trying to suppress competition from legitimate resales.

Contributory Trademark Infringement

The flagship case articulating an online standard for contributory trademark infringement is the 1999 Ninth Circuit Lockheed v. NSI case. This court was squarely presented with the issue of the Second Circuit’s standard, and the court punted. Instead, the court said that both eBay and Tiffany had agreed to use the Inwood standard (referencing a 1982 Supreme Court opinion, the leading offline case for contributory trademark infringement). The court recaps the standard: “there are two ways in which a defendant may become contributorially liable for the infringing conduct of another: first, if the service provider ‘intentionally induces another to infringe a trademark,’ and second, if the service provider ‘continues to supply its [service] to one whom it knows or has reason to know is engaging in trademark infringement.’” Inducement was not an issue here.

eBay followed a notice-and-takedown scheme, so actual knowledge was not at issue either. eBay’s generalized knowledge of counterfeiting activities on the site was insufficient; the court required “[s]ome contemporary knowledge of which particular listings are infringing or will infringe in the future.” Tiffany’s demand letters weren’t specific enough, and its specific requests were honored. Therefore, eBay lacked the requisite scienter.

Unfortunately, the court did not stop there. Instead, the court brought up the possibility of “willful blindness.” The court articulated a new and troublesome legal standard: “When [a service provider] has reason to suspect that users of its service are infringing a protected mark, it may not shield itself from learning of the particular infringing transactions by looking the other way.”

What does this mean??? What lays the foundation for a service provider to have “reason to suspect” that users are infringing? Does a C&D letter do that? The answer seems to be no; Tiffany sent those. Does proactive filtering confer that foundation? Perhaps, because eBay did undertake some extra filtering efforts for Tiffany. In FN 14 the court says “contributory liability may arise where a defendant is (as was eBay here) made aware that there was infringement on its site but (unlike eBay here) ignored that fact.”

It’s great for eBay that it didn’t ignore the fact, but what exactly did eBay do right (other than win at trial court)? I’m not sure, and that’s telling for the weaknesses in the court’s language and logic. As a result, I expect plaintiffs to get frisky with this “willful blindness” toy and start asserting that defendants had “reason to suspect” user infringement and “ignored that fact.”

Sadly for Tiffany, it won’t get the benefit of this loose language. The court concludes its discussion on willful blindness by saying “eBay appears to concede that it knew as a general matter that counterfeit Tiffany products were listed and sold through its website….Without more, however, this knowledge is insufficient to trigger liability under Inwood. The district court found, after careful consideration, that eBay was not willfully blind to the counterfeit sales….That finding is not clearly erroneous. eBay did not ignore the information it was given about counterfeit sales on its website.”


The court seems to say that there cannot be blurring or tarnishment when the defendant makes a nominative use. The court says simply “There is no second mark or product at issue here to blur with or to tarnish ‘Tiffany.’” This is interesting because the post-TDRA statute expressly provides a defense for nominative use, but the court doesn’t rely on it. Instead, it seems to treat nominative use as a definitional requirement–if the defendant is making a nominative use, tarnishment and blurring should be categorically impossible and defendants should win on a 12(b)(6) motion to dismiss. But if that’s the case, why have the statutory defense? (Don’t get me started on the “commercial use” infinite loop in the dilution statute).

To the extent counterfeiting causes dilution, eBay didn’t sell the goods itself, so it did not commit that dilution. The court mucks its words about contributory dilution; what it should have said is that there is no such thing.

False Advertising

The court describes eBay’s advertising at issue:

eBay advertised the sale of Tiffany goods on its website in various ways. Among other things, eBay provided hyperlinks to “Tiffany,” “Tiffany & Co. under $150,” “Tiffany & Co.,” “Tiffany Rings,” and “Tiffany & Co. under $50.”…eBay also purchased advertising space on search engines, in some instances providing a link to eBay’s site and exhorting the reader to “Find tiffany items at low prices.”

The court says that these statements are not literally false because genuine Tiffany merchandise was available on eBay, “[b]ut we are unable to affirm on the record before us the district court’s further conclusion that eBay’s advertisements were not ‘likely to mislead or confuse consumers.’” The court effectively rejected the district court’s use of nominative use as a categorical defense to false advertising claims. I wasn’t too surprised to see this; in my post about the district court ruling, “I fear the court may have cut some corners here.”

What about 47 USC 230? eBay has used the immunity to defend its ad copy from non-IP claims to the extent the ads are rendered untrue by third parties (see the uncited Mazur case). The court does not directly address 230, but it would not be a fan of such a defense: “eBay did affirmatively advertise the goods sold through its site as Tiffany merchandise. The law requires us to hold eBay accountable for the words that it chose insofar as they misled or confused consumers.”

Then, the court takes another weird detour. eBay appears to be free to advertise that it vends legitimate Tiffany goods, but it can’t mislead consumers into thinking that all Tiffany goods on eBay are legitimate. How is eBay supposed to strike that balance? The court says: “An online advertiser such as eBay need not cease its advertisements for a kind of goods only because it knows that not all of those goods are authentic. A disclaimer might suffice. But the law prohibits an advertisement that implies that all of the goods offered on a defendant’s website are genuine when in fact, as here, a sizeable proportion of them are not.”

Wait a minute. What kind of disclaimer is the court thinking of here? Maybe “Please come to eBay for Tiffany goods. Disclaimer: we have lots of fakes on our site.” Putting aside the implications for eBay’s scienter, how is eBay supposed to fit such a disclaimer into the tight space constraints of an AdWords ad? This reminds me a little of the FDA’s threats that it expected pharma companies to make side effects disclosures in AdWords copy…NOT POSSIBLE.

And what constitutes a “sizeable proportion”? The parties did not agree on the exact amount of counterfeiting taking place, and Tiffany’s expert report was partially discredited by the trial judge. Does this mean we’re back to taking snapshots of infringing activity and arbitrarily picking a percentage as too high? I thought we got past that with the Grokster Supreme Court opinion.

Rebecca has useful things to say about the false advertising discussion.

Implications of this Ruling

Between this ruling and the Akanoc ruling (and perhaps the Google ECJ opinion, if you look internationally), I think it’s pretty clear now that trademarks are subject to a notice-and-takedown rule. We don’t have a DMTA to complement the DMCA’s notice-and-takedown procedure, and perhaps we would benefit from some statutory clarity about the precise mechanics/specifications of the notice and takedown. But even without a DMTA, I think we’ve arrived at the same basic place. From my perspective, this is not inherently good or bad. I guess I’ve always thought we’d get here eventually.

I continue to believe that courts reward service providers who do more than statutorily required for IP owners. For example, in FN 16, the court favorably notes that “eBay’s efforts to combat counterfeiting far exceeded the efforts made by the defendants in” the flea market cases. I continue to encounter pockets of folks who believe that undertaking any voluntary efforts above the minimum makes the site a guarantor against bad activity. I think that’s very outdated thinking. When I asked above what eBay did right to avoid being willfully blind, the answer is that they were proactive about working with IP owners. This case shows that those good faith efforts can be rewarded in court. UPDATE: Greg Lastowka explores this point more.

Having said that, I remain concerned that efforts to cater to IP owners’ concerns can create a competitive barrier to entry. Not every site is sufficiently capitalized to undertake the same efforts as eBay. I hope courts will be circumspect about letting the law effectively reduce service provider competition from new entrants.

It remains to be seen what result this ruling will have on the keyword advertising cases. Through its broad reading of nominative use, it’s possible the Second Circuit implicitly undercut some of the lawsuits against Google.

Finally, junky false advertising claims are so ubiquitous that we often don’t give them much respect. However, here’s a good example where the false advertising claim was more potent than the trademark claim. I believe false advertising will remain a hot area of legal practice.