Following its successful appearance before the Court of Justice of the European Communities (ECJ) earlier this year, Google recently announced sweeping changes to its AdWords policies in the European Union. As a result, Google now permits the acquisition of third-party trademarks as AdWords in the European Union, and also allows some advertisers to utilise third-party trademarks in advertisement text without the permission of the brand holder. Google has stated that "in response to a complaint from a trademark owner, we will do a limited investigation as to whether a trademarked keyword in combination with particular ad text is confusing as to the origin of the advertised goods and services. If we find that it is, we will remove the specific ad that is the subject of the complaint". This language, closely tracking that used by the ECJ in its *AdWords* decision, obviously leaves a lot of room for interpretation. Nonetheless, these changes bring Google policy in the European Union into closer alignment with the policies that have been in place in the United States, Canada, the United Kingdom and Ireland since last year.

So what does this mean for brand holders in the United States and in Europe? While some European trademark owners have understandably expressed trepidation about the new policy (and while it seems that every trademark policy announcement by Google is met with a visceral reaction by some in the IP bar on both sides of the pond), on balance, the relative uniformity should be welcomed - particularly among those brand owners that are attempting to coordinate strategies and policies across many jurisdictions.

Indeed, although the United States and Europe utilise different nomenclature, the underlying substantive trademark law is quite similar. Thus, these developments with regard to Google policy should lead to a more efficient analysis of the relevant issues. Furthermore, challenges under the new system should become more streamlined, requiring fewer resources to be invested and removing incentives for forum shopping.

European trademark owners can learn from the US example, where these policies have been in place long enough for some well-documented legal challenges to be brought.

For there to be infringement under US federal law, the allegedly infringing use must:

  • be a "use in commerce"; and
  • give rise to a "likelihood of confusion".

Taken together, this framework has shaped US jurisprudence on keyword-triggered advertising.

The 2009 US Second Circuit Court of Appeals decision in *Rescuecom v Google* (562 F 3d 123) eliminated a split of authority among US appeals courts and more firmly established that use of a third-party trademark in Google's AdWords programme, and its KeyWords Suggestions Tool programme, does constitute a potentially actionable "use in commerce".

When it was issued, the *Rescuecom* decision came as something of a surprise, insofar as it seemed to cut against the Second Circuit's more established precedent, such as in *1-800-Contacts v* (414 F 3d 400). In any event, the net effect of *Rescuecom* seems to be that it brought at least the US federal courts into consensus regarding the classification of keyword-triggered advertising as use in commerce in the search engine context.

The battleground, then, has shifted to whether the sale of trademarks as search terms in programmes such as Google's AdWords creates a likelihood of confusion. Although the United States has a well-developed body of case law regarding the likelihood of confusion analysis, that doctrine has not been fully tested in the online keyword-triggered advertising context. Several courts have suggested that a likelihood of confusion will be difficult to prove - at least where the trademark is used solely as a keyword and not in the actual advertisement text.

Indeed, in *JG Wentworth, SSC Ltd Partnership v Settlement Funding LLC* ((85 USPQ 2d 1780 (EDPa, 2007)) one US district court held that if the mark does not appear in the advert text, there can be no likelihood of confusion. More recently, in *Rosetta Stone Ltd v Google Inc*, another US district court held that "no reasonable trier of fact could find that Google's practice of auctioning Rosetta Stone's trademarks as keyword triggers to third party advertisers creates a likelihood of confusion as to the source and origin of Rosetta Stone's products".

Another area of potential future litigation in this arena in the United States is the extent of the interaction between the concept of likelihood of confusion and the US nominative fair use doctrine.

Nominative fair use is a carve-out from the infringement analysis for those cases in which the alleged infringer is using the brand owner's trademark to refer to the brand owner's goods or services. The policy backbone justifying the existence of the doctrine is straightforward: how can the infringer cause confusion when the mark is being used accurately, to refer to the brand owner's goods? The seminal case of nominative fair use in the United States, *New Kids on the Block v News America Publishing* (971 F2d 308, 308 (9th Cir 1992)), involved a survey about teen band New Kids on the Block. The court held that because there was simply no way to ask questions about the band without using the referencing "New Kids on the Block", the publisher's use was fair and not an infringement. In practice, however, at times the application of the nominative fair use doctrine can be quite tricky and nuanced.

One important facet of the doctrine is that in order to qualify for the protection afforded by the nominative fair use doctrine, a party can use no more of the trademark than is necessary. For example, in a number of cases outside the internet context, courts have held that independent repair shops that work on Volkswagen cars could in fact advertise that they repair VOLKSWAGEN-branded vehicles. However, the courts have also held that the repair shops could not make use of the Volkswagen logos or other trademarks. In the internet context, however, a US court has held that a former Playboy Playmate could use the marks PLAYBOY and PLAYMATE to promote her website, including within the site's metatags, as the terms were accurately used to describe her.

The application of the nominative fair use doctrine to keyword advertising, however, has not yet been fully vetted in the US. For example, if an independent Volkswagen repair shop purchased the keyword "Volkswagen" in the Google AdWords programme, it seems that the repair shop would have an equally strong nominative fair use argument as in the offline context. The influential US District Court for the Southern District of New York found there was a nominative fair use under analogous facts in the much-publicised 2008 decision in *Tiffany Inc v Ebay Inc* (Case 04 Civ 4607 (RJS), July 14 2008). However, it is less clear whether, if the repair shop were to use the term VOLKSWAGEN in the sponsored link text, this would still qualify as a nominative fair use.

In a similar vein, the application of the comparative advertising doctrine in the keyword advertising context is also not yet certain. Paradigmatically, in comparative advertising, a company uses its competitor's trademark to assert that its own product or service is better or less costly (eg, "Whoppers are tastier than Big Macs!"). Like nominative fair uses, such a comparison is generally not considered infringing under US federal law. However, if Burger King purchased the keyword "Big Mac" and then placed the same "Whoppers are tastier than Big Macs!" text in a sponsored link, the interplay between the existing legal framework governing comparative advertising and the likelihood of confusion analysis would become significant.

Despite the many ambiguities in US law in need of adjudication, there appears to have been a drop in the number of infringement cases filed against Google since *Rescuecom* established that AdWords are "use in commerce" for purposes of US trademark law. There are a number of potential reasons for this.

On a practical level, one plausible explanation is that the economic downturn in the United States has caused a drop in all litigation. Another reason is that, as discussed above, it is very difficult to prove a likelihood of confusion in the keyword advertising context, and the courts have shown their reluctance to find a likelihood of confusion, particularly where the trademark at issue does not appear in the text of the sponsored link.

However, further tests of the limits of these new axioms in both the United States and Europe are inevitable. Google may find itself defending its own determinations made in furtherance of its European policy of investigating and removing ad text that is purportedly "confusing as to the origin of the advertised goods and services". However, signs suggest that advertisers, rather than Google itself, may more often be in the cross-hairs of this next round of litigation.

Thus, there will undoubtedly continue to be US and EU litigation concerning Google's policy of permitting third parties to employ third-party trademarks as Google AdWords. Nonetheless, while many brand holders may be apprehensive about Google's recent policy changes, the relative alignment of Google's European with US policies represents a step forward. Whatever your perspective on the legalities of Google's policies and practices, the efficiencies created by a consistent policy across jurisdictions should be welcomed.

That said, open questions do remain regarding, among other things, the application of the US likelihood of confusion analysis and US law concerning nominative fair use and comparative advertising to keyword advertising, particularly where the third-party trademark appears in the advertisement text. This means that brand owners and those who advise them would be well served to continue to stay closely informed about this rapidly changing area of trademark law, and about Google's continued policy development.