The closely watched battle over the use of trademarks as keywords for purpose of triggering advertisements on Google’s search result pages (AdWords) reached high peak today with the release of the European Court of Justice’s ruling on the French cases. In what appears to be a resounding win for Google, the ECJ managed to avoid some of the critical questions in a decision that, in fact, projects little new light on the multibillion dollars question: Is AdWording that involves marks as keywords legal in Europe?
The French courts, in which some highly publicized keywording disputes were being litigated, have referred to the ECJ a series of questions relating to the practice of using protected trademarks to trigger advertisements placed by third parties without the permission of the markholders.
The key legal provisions are Articles 5(1)(a)-(b) and 5(2) of the European Trademark Directive (1988) (89/104/EEC). (In the meantime, a more recent trademark directive (2008) went into effect, which is inapplicable to the disputes at hand, but which contains essentially similar provisions concerning the “[r]ights conferred by a trade mark” (Art. 5(1)-(2) of the 2008 Directive)).
Art. 5(1)(a) to the Trademark Directive governs use of signs identical to the protected mark in connection with goods or services identical or similar to the goods or services for which the mark is registered, whereas Art. 5(1)(b) covers use of identical or similar signs in a manner that causes likelihood of confusion. Art. 5(2), on its part, covers dilution situations, in which signs that are identical or similar to a reputable marks (aka “famous marks”) are used in relation to goods/services that are NOT similar to those for which the mark is registered.
The analysis of the ECJ proceeded on three levels. FIRST, according to E.U. law, claims against advertisers and against Google, in all three subsets of infringing activities, must show that the accused party has used the protected mark “in the course of trade.” In the context of identical signs, the threshold is very low: The use-in-the-course-of-trade test is satisfied “where [use] occurs in the context of commercial activity with a view to economic advantage and not as a private matter.” (¶50). Accordingly, the ECJ ruled that AdWords advertisers that use keywords identical to protected marks in order to trigger their ads certainly qualify, apparently including AdWords that do not exhibit the protected mark in any way (so-called “invisible use”).
However, the Court found that Google does not use the mark in the course of trade, even though it clearly draws economic advantage from AdWording practices that do. It is important to note that pulling Google off the hook already at this stage insulates it from allegations covered by all three subsets of illicit activities regulated under Art. 5(1) and Art. (2). How did the Court reach this conclusion? Unfortunately, the reasoning is quite murky (¶¶ 53-59). The Court seems to draw a distinction between Google and advertisers on the theory that Google, as opposed to advertisers, does not send its OWN message to consumers. (I think that the intuition behind this reasoning is correct, but, by the same token, the Court should have legitimized messages from advertisers to the public that do not mention the mark to consumers in any way). Interestingly, the Court characterized Google’s role as merely “storing” the keywords at question, which, as the Court concludes, does not qualify as use in the course of trade, but, as discussed below, is relevant in the context of the European safe harbor provisions.
On the SECOND level, an illicit use of an identical sign (Art. 5(1)(a)) requires use “in relation to goods or services.” Here, the ECJ applied an expansive standard that clearly covers also “invisible” use of the mark: Even in case that the advertiser makes no reference to the mark in the AdWord or otherwise while attempting to promote its own goods and services, use of the protected mark occurs simply by virtue of booking the keyword with Google. The Court explicitly rejected Google’s counter argument on this crucial point (¶¶ 64-73), thereby making a quite remarkable observation:
It is also clear that in most cases an internet user entering the name of a trade mark as a search term is looking for information or offers on the goods or services covered by that trade mark. Accordingly, when advertising links to sites offering goods or services of competitors of the proprietor of that mark are displayed beside or above the natural results of the search, the internet user may, if he does not immediately disregard those links as being irrelevant and does not confuse them with those of the proprietor of the mark, perceive those advertising links as offering an alternative to the goods or services of the trade mark proprietor.
Hmm. I am not aware of compelling empirical studies on this question (and I am not sure the Court has seen any either). Ads, by definition, relate in some way to goods or services, and most naturally, to goods/services of the same entity that pays for them. The presumption that ads appearing under the “sponsored links” rubric on the right side of the natural search results list always (or even often) offer goods/services “alternative” to goods/services of an entity that happens to hold trademark rights in the search term strikes me as speculative.
On the THIRD level, to constitute trademark infringement, the use must also have an adverse effect on the protected functions of the mark. In this respect, the ECJ mentioned two such protected functions. The first, more traditional function concerns the indication of origin. Can an “invisible use” of a mark that results in the placement of ads under the rubric of “sponsored links” harm this function? The analysis of the ECJ is inconclusive (¶¶82-90). The Court considers it a matter for national courts to resolve on a case-by-case basis. That said, the Court still added two important instructions to national courts in this regard:
In the case where a third party’s [advertiser’s] ad suggests that there is an economic link between that third party and the proprietor of the trade mark, the conclusion must be that there is an adverse effect on the function of indicating origin.
And:
In the case where the ad, while not suggesting the existence of an economic link, is vague to such an extent on the origin of the goods or services at issue that normally informed and reasonably attentive internet users are unable to determine, on the basis of the advertising link and the commercial message attached thereto, whether the advertiser is a third party vis-à-vis the proprietor of the trade mark or, on the contrary, economically linked to that proprietor, the conclusion must also be that there is an adverse effect on that function of the trade mark.
This latter instruction – the requirement to find “vagueness” concerning origin from the perspective of “normally informed and reasonably attentive internet users” is perhaps the crux – and the anticlimax – of the whole decision. The Court practically throws back to national courts the most controversial and difficult question, precisely the issue in which they have been desperate for some more guidance. This is the place where the ECJ should have come up with some more concrete instructions, but came up with a vague test of vagueness instead.
The ECJ also mentioned briefly the so-called “advertising function” of the mark, namely, the interest of a markholder in “using its mark for advertising purposes designed to inform and persuade consumers.” Here the Court offers something that comes closer to a bright-line rule: It held that possible “repercussions” on the advertising use caused by AdWords do not in themselves constitute adverse effect on the advertising function of the trademark (¶95). Further, “it must be concluded that use of a sign identical with another person’s trade mark in a referencing service such as that at issue in the cases in the main proceedings is not liable to have an adverse effect on the advertising function of the trade mark” (¶98).
The Court continued by evaluating liability under Art. 5(2) (dilution). It held that in case that an advertiser applies AdWords in order to promote the sale of imitations (e.g., of Louis Vuitton’s goods), the advertiser is liable for dilutive use, yet Google is not – for the reasons eluded to above: Merely “storing” the keyword does not amount generally to using it in the meaning of the Directive, so it cannot specifically violate its dilution provision.
Alongside the Trademark Directive, also the E-Commerce Directive (2000), and specifically, its safe harbor provisions, received some limited treatment. Can Google enjoy the immunity provisions concerning “storage” of infringing information in accordance with Art. 14 of the E-Commerce Directive?
Well, also here the Court answers with a categorical 'it depends!' Namely, it delegates the task of assessing the role of Google on a case-by-case basis to national courts. More concretely, national courts are called to resolve whether Google’s operation qualifies it as a passive conduit, or, in other words, whether Google “is neutral in the sense that its conduct is merely technical, automatic and passive, pointing to a lack of knowledge or control of the data which it stores.” (In terms of Art. 14(1)(a), it is a condition to immunity that “the provider does not have actual knowledge of illegal activity or information and, as regards claims for damages, is not aware of facts or circumstances from which the illegal activity or information is apparent.” The ECJ did not add much substance on top of the plain language of the Directive on this point.)
Bottom line: In the aftermath of this ruling, much remained unsettled. I think that the unequivocal language concerning the satisfaction of the use “in relation to goods or services” requirement should make European advertisers who use keywords identical to marks very, very concerned. It might even kill this practice in certain European countries, hardly a reason for celebration in Google’s quarters, to be sure. At the same time, the core legal question, whether invisible use of marks in AdWords constitutes trademark violation by advertisers, did not receive a satisfactory answer. Finally, though Google can now rest assured that direct liability for trademark infringements committed by advertisers will not attach to its own activity as a platform provider, many questions concerning secondary liability, both for trademark infringement and for other violations (such as unfair competition) remain wide open.