D. Expansion of Inwood Standard to “Non-Product” Cases: The “Direct Control and Monitoring” Test: 2. Internet Activity: (c) Search Engine Companies
When courts consider search engine company liability for trademark infringement, whether direct or indirect, they typically focus on the sale of “trademarked” keywords, or search terms, to third parties for use in advertising and directing internet traffic to websites that compete with the trademark owner. See e.g., Rescuecom Corp. v. Google Inc., 562 F.3d 123 (2d Cir. 2009) (search engine company’s use of plaintiff’s mark constituted a use in commerce), vacating and remanding, 456 F.Supp.2d 393 (N.D.N.Y 2006); and see, Google Inc. v. American Blind & Wallpaper Factory Inc., 2005 WL 832398 (N.D. Cal. 2005); Government Employees Ins. Co. v Google, Inc., 330 F.Supp.2d 700, 704-705 (E.D. Va. 2004) (“GEICO”), discussed below. If such use were itself infringing, contributory liability would arise out of the search engine company’s role in selling the infringing keywords to third parties.
Contributory liability under Inwood may also extend to a search engine company, it has been argued, if the company knowingly continued to supply its services to an infringing third party. Rosetta Stone Ltd. v. Google Inc., 2010 WL 3063152, *13 – *14 (E.D. Va.).This was the basis for liability urged unsuccessfully by the luxury jewelry company Tiffany in its suit against the online sale site eBay. See Tiffany v. eBay, 600 F.3d 93 (2d Cir. 2010), affirming in part and remanding in part, 576 F.Supp.2d 463 (S.D.N.Y. 2008), cert denied, 131 S.Ct. 647 (2010). The Rosetta Stone court drew extensive comparisons between the two cases in reaching its decision. (Tiffany is discussed in detail in Section II.B.3.(a))
In Rosetta Stone, where the plaintiff language-learning company argued both theories of liability against the search engine Google, the court declined to find contributory liability under either approach, ruling in favor of Google on a motion for summary judgment on the claim. Rosetta Stone, supra at *14 (E.D. Va.). In doing so, however, the court limited its ruling to “advertisers selling counterfeit Rosetta Stone products,” leaving the precedential value of the opinion unclear. Id. at *13.
The dispute in Rosetta Stone arose out of Google’s “AdWords Propram,” an “auction-style advertising program that displays advertisements to users of Google’s search engine in the form of Sponsored Links.” Rosetta Stone, supra at *3. With the AdWords Program, advertisers can use Google to cause a Sponsored Link to their website to appear whenever a user searches on Google for certain keywords. Prospective advertisers can select keywords they choose themselves or from a list provided by Google. Id. Lists of keywords provided by Google are “filtered” by Google to remove “trademarked” terms for which Google has received a complaint. Id.
At the time of the litigation, Google’s AdWords policy expressly allowed a given keyword to be used both as a trigger to a Sponsored Link advertisement and as part of the advertisement itself. Rosetta Stone, supra at *4. More fundamentally, Google’s policy permitted not only the brand owner and its authorized licensees, but also other advertisers to include another’s trademark-protected term in their advertisement text under certain, non-infringing conditions. See Id. To address fraud and counterfeiting associated with its AdWords Program, Google had created a “Trust and Safety Team.” Rosetta Stone, supra at *4. The Trust and Safety Team responded to “notices of counterfeit advertisements on Google’s website and [took] down any advertisements confirmed to violate its AdWords Program.” Id.
Notwithstanding Google’s internal policing, Rosetta Stone asserted that some advertisers were able to beat the system and create Sponsored Links that “deceive[d] and misdirect[ted] Google’s users to websites that [sold] counterfeit Rosetta Stone products or suggest[ed] to consumers a connection to Rosetta Stone that [did] not exist.” Id. It therefore sued Google, alleging, in addition to direct infringement, both contributory and vicarious trademark infringement. (The vicarious liability discussion appears in Section III.)
In support of the contributory liability claim, Rosetta Stone argued the two theories of infringement outlined above, tracking the two-prong test under Inwood. (See Section II.B. for a full discussion of the Inwood test.) Thus, as to intentional inducement, Rosetta Stone contended that Google’s practice of including brand names as suggested keywords “directly induce[d] advertisers to infringe on Rosetta Stone’s marks.” Rosetta Stone, supra at *13.
Secondly, Rosetta Stone argued that “by allowing counterfeiters to open AdWords accounts and bid on Rosetta Stone Marks, despite receiving notice of their counterfeit status, Google [was] supplying a service to those it [knew] or ha[d] reason to know [were] engaging in trademark infringement.” Rosetta Stone, supra. To demonstrate Google’s knowledge of ongoing infringement, Rosetta Stone pointed to an admission by Google in its Registration Statement filed with the Securities and Exchange Commission acknowledging that as a result of its policy allowing the purchase of trademark-protected keywords, it could be subject to more trademark infringement lawsuits. Id. at *4 and *13. It pointed furthermore to “approximately 200 instances of Sponsored Links advertising counterfeit Rosetta Stone products[,]” and asserted that even after being notified about the counterfeit websites involved, “Google continued to allow Sponsored Links for other websites by these same advertisers to use the Rosetta Stone Marks as keyword triggers and in the text of their Sponsored Link advertisements.” Id. at *13.
The court rejected both of Rosetta Stone’s arguments. Applying Inwood, it reiterated that
[t]o prevail on a contributory trademark infringement claim, a plaintiff must show that the defendant “intentionally induces another to infringe a trademark, or  continues to supply its product to one whom it knows or has reason to know is engaging in trademark infringement…”
Rosetta Stone, supra, citing Inwood Labs., Inc.v. Ives Labs., Inc., 456 U.S. 844, 854 (1982). It recognized further the application of the Inwood test beyond manufacturers and distributors, to service providers who “exercise sufficient control over the infringing product.” Id., citing Tiffany v. eBay, 600 F.3d 93, 104 (2d Cir. 2010) and Lockheed Martin Corp. v. Network Solutions, Inc., 194 F.3d 980, 984 (9th Cir. 1999).
As to intentional inducement, the court disagreed with Rosetta Stone’s stance on Google’s program regarding keywords, reasoning that “the mere existence of a tool that assists advertisers in optimizing their advertisements does not, in itself, indicate intent to induce infringement.” Rosetta Stone, supra at *14. Because Google’s policy could be attributed to “good business practice” rather than intent to induce infringement, the court declined to extend contributory liability based on this theory, concluding that “[a] desire for economic gain alone does not translate into contributory trademark infringement.” Id. Note that the court did not elucidate at this point whether it referred to the infringement arising out of the use of the trademark-protected keywords, or that associated with websites selling counterfeit Rosetta Stone products. See Id.
The court drew heavily on the Second Circuit’s reasoning in Tiffany v. eBay to reject Rosetta Stone’s second argument, finding that it failed to show that Google knew or had reason to know that it was supplying its services to parties engaging in trademark infringement. Rosetta Stone, supra at *14, citing 600 F.3d 93, at 107-109 (2d Cir. 2010), affirming in part and remanding in part, 576 F.Supp.2d 463 (S.D.N.Y. 2008), discussed in detail in Section II.B.3.(a). In Tiffany, “eBay’s generalized knowledge of infringement” on its website was insufficient to impose contributory liability on eBay. Id. Compared to eBay, the court reasoned, Google had been presented with even fewer notices of infringement. It therefore declined to find the “specific” knowledge required by the court in Tiffany. Id. Additionally, the court noted with approval Google’s own internal policing procedures regarding advertisements for counterfeit goods, acknowledging that “Google [had] worked closely with law enforcement and brand owners to combat counterfeiting because it [knew] that those advertisements [could] create a bad experience for web users, who Google ultimately reli[ed] on for its business.” Id. Comparing the two cases further, the court observed that like eBay, Google was in no position to determine which websites were selling counterfeit Rosetta Stone products. Rosetta Stone thus failed to convince the court to extend contributory liability based on the “continues to supply” argument under Inwood. Id.
In finding that Rosetta Stone had not met its burden of showing that summary judgment was proper as to the contributory trademark infringement claim, the court’s language left unclear what its conclusions were regarding Google’s liability for the sale of trademark-protected keywords to competitors who were not necessarily counterfeiters. See Rosetta Stone, supra at *13 -*14. Specifically, it found that “no reasonable trier of fact could find that Google intentionally induce[d] or knowingly continue[d] to permit third party advertisers selling counterfeit Rosetta Stone products to use the Rosetta Stone Marks in their Sponsored Link titles and advertisement text.” Id. at *13. If indeed the court intended to limit its holding to direct trademark infringement arising out of counterfeit websites, then it is not obvious how or whether its decision would apply to the scenarios presented in the cases of GEICO and American Blind, infra. In these two earlier cases, the courts left open the possibility that search engine operators who use third parties’ trademarks might, in appropriate circumstances, bear contributory liability for trademark infringement arising out of those advertisements. Google Inc. v. American Blind & Wallpaper Factory Inc., 2005 WL 832398 (N.D. Cal. 2005); Government Employees Ins. Co. v Google, Inc., 330 F.Supp.2d 700, 704-705 (E.D. Va. 2004) .
GEICO, supra, was a case of first impression in which the insurance company GEICO sued Google and Overture (an Internet marketing company) for using GEICO’s trademarks to sell advertising. GEICO alleged, inter alia, that the defendants were contributorily liable for the paid advertisements which contained GEICO’s marks in their text and were generated by customers who selected those terms when conducting a search. The defendants moved to dismiss the case for failure to state a claim, but the court refused, permitting the case to go forward on the issue of contributory liability.
Specifically, GEICO claimed that the defendants’ contributory liability arose “when the advertisers themselves [made] use of the GEICO marks by incorporating them into the advertisements, which [were] likely to deceive customers into believing that the advertisers provide[d] accurate information about GEICO.” GEICO, supra 704. It further alleged that the defendants exercised significant control over the content of those advertisements. Id. The court found that these allegations were sufficient to support a claim of contributory infringement. See id.
Furthermore, regarding Overture, GEICO had alleged that that company “encourage[d] advertisers to bid on trademarked words, and monitor[ed] and control[led] the allegedly infringing third-party advertisements.” GEICO, supra at 705. Apparently referring to its internal trademark policing policy, Overture argued that “its monitoring [was] intended to prevent, not encourage, trademark infringement.” Id. This argument, the court noted, raised a disputed fact that could not appropriately be resolved by a motion to dismiss. Id. Moreover, and of potentially critical importance, the claim by GEICO that Overture monitored and controlled the third-party advertisements was sufficient to plead the actual or constructive knowledge required to allege contributory infringement. Id. (citations omitted). For a discussion of whether the “knowledge” requirement ought to be subsumed properly by the “monitoring and control” requirement, as the court appears to suggest here, see discussion at supra.
In a similar case, Google Inc. v. American Blind & Wallpaper Factory Inc., 2005 WL 832398 (N.D. Cal. 2005), Google sought a declaratory judgment from the court that its “AdWords” advertising program, described above, did not infringe on the defendant’s trademarks. The defendant, American Blind, made a motion to dismiss, which was denied. It then answered and responded with counterclaims for both direct and indirect infringement of its trademarks against not only Google, but also against various other companies that operate web sites that included an Internet search engine. Those companies, including “Ask Jeeves,” Earthlink, AOL, Netscape and Compuserve, paid Google to access its Web-searching platform, thus yielding similar results to those displayed by Google. See Google, supra at *3. The defendants then moved to dismiss American Blind’s counterclaims.
American Blind’s counterclaims for contributory trademark infringement centered on Google’s “Adwords” program. Through this program, American Blind contended, Google had sold to American Blind’s competitors, “keywords” that included in some form the American Blind trademarks, including AMERICAN BLIND, AMERICAN BLINDS, and AMERICANBLINDS.COM, which sales had proceeded over its objection. Id. at *2. American Blind specifically fingered Google’s “Adwords Keyword Suggestions” feature, through which Google actively encouraged its competitors to purchase as keywords not only the American Blind trademarks but also “every conceivable” variation thereon. See id. Thus “an advertiser who is considering purchasing the keyword “American Blind” is encouraged also to purchase the keywords “American blinds,” “American blinds and wallpaper,” and “american blinds and wallpaper factory,” among others.” Id. It noted that Google even had a name for this practice, in its case, “the American Blind optimization campaign” Id. It further contended that although Google had the capacity to block the purchase of its marks, it had, pursuant to a major shift in its trademark policy, made a conscious decision “not to disable Sponsored Links when advertisers have purchased keyword triggers that are trademarked terms.” Id.
Based on the above practices, American Blind alleged that Google had induced its competitors to purchase the American Blind marks as keywords and had refused to block its competitors’ advertisements resulting from searches for those marks, notwithstanding their knowledge that such advertisements infringed on the marks. See id. at *7, n.27. It bears noting, in this regard, that American Blinds apparently argued for a direct application of Inwood, i.e., inducement, as opposed to an application of Inwood in the non-product context. Google argued, in response, that American Blind had failed to allege the requisite “use” of its marks, in this case by the defendant’s advertisers themselves. See id. at *7. Specifically, Google contended that “because their advertisers are alleged to use the American Blind Marks only as a trigger for the display of their advertisements and not as an identification of the source of their products, they [did] not engage in trademark ‘use’ of the American Blind Marks[.]” Such a failure to allege a direct infringement of its trademark rights, Google argued, defeated its contributory liability claim. Id.
The court disagreed, rejecting Google’s motion to dismiss and allowing the contributory trademark infringement claim to go forward. Drawing on its earlier analysis regarding American Blind’s claims of direct trademark infringement, see id. at *4 – *7, the court was not convinced, given the unsettled nature of the law, that the purchase of trademarks as keywords did not constitute trademark “use” by Google’s advertisers. Id. at *7. Compare Google v. American Blind with 1-800 Contacts v. Lens.com, 755 F.Supp.2d 1151, 1174 (D. Utah 2010), where the court held that “the mere purchase of a trademark as a keyword cannot alone result in consumer confusion” and therefore was non-infringing. Citing the liberal legal standard regarding motions to dismiss, which are viewed by the courts with disfavor, see Id. at 3, the court concluded that American Blind had made sufficient allegations regarding both direct and contributory liability that it did not appear “beyond doubt” that American Blind could prove no set of facts in support of its claims that would entitle it to relief. Id. at *7. For further discussion regarding the extent to which a plaintiff must demonstrate direct trademark infringement in order to bring an action for contributory trademark infringement, see discussion at II.B.5.