After several years of litigation over the use of trademarks in paid search advertisements, or AdWords, Google recently introduced new policies protecting trademark owners. Google now requires written permission from the trademark owner in order to use a trademark term in the text of a paid search ad—even if the advertiser is a legitimate reseller of the trademarked product. This is a curious development considering that it departs from Google’s past practice of being less aggressive with regard to trademark issues. For example, Google long ago abandoned a policy of permitting trademark owners to block bidding on trademarks used as ad words. The explanation for this change appears to be linked to a rare defeat for Google in a lawsuit relating to its AdWords advertising service.
Back in 2009, Rosetta Stone sued Google for trademark infringement, contributory infringement, and dilution based on Google’s practice of allowing acknowledged counterfeiters of Rosetta Stone software to purchase sponsored search results using Rosetta Stone’s name. Although the suit was initially dismissed by a federal court in 2010, in April 2012, the United States Court of Appeals for the 4th Circuit reinstated Rosetta Stone’s claims. The 4th Circuit found that there was a genuine question of fact as to whether web users would be confused about the source of sponsored search results purchased by acknowledged counterfeiters of Rosetta Stone software. The court’s conclusion was based in part on internal Google documents that showed that consumers are often confused as to the meaning of sponsored search results. Moreover, the evidence in the case showed Rosetta Stone repeatedly notified of Google of paid search ads for counterfeit software—over 190 times in a seven month span.
The Rosetta Stone case is unusual in that it may be the first time that a trademark-related case involving AdWords appeared to be headed to trial. (In the US, that is—Google has lost a number of trademark cases in Europe and in fact has different AdWords policies there as a result). Not surprisingly, the case settled. No specific terms of the settlement have been disclosed, but in a somewhat terse joint statement, Google and Rosetta Stone announced their intention to “collaborate to combat online ads for counterfeit goods and prevent the misuse and abuse of trademarks on the Internet.” Shortly thereafter, Google amended its Ad Words policies. Given the language of the joint statement, and the timing of the subsequent change in Ad Words policies, it seems likely that the Rosetta Stone case is responsible for the new requirement that trademark owners provide permission for the use of trademarks in paid search ad copy.
Beyond the change to Google’s policies, the case is significant going forward as well. It suggests that Google may be held responsible in instances in which it has knowledge that the ads create confusion and/or point to counterfeiters’ web sites. This marks a new development in the area of paid search infringement analysis that bears further watching.
Showing posts with label Trademarks. Show all posts
Showing posts with label Trademarks. Show all posts
Thursday, May 23, 2013
Tuesday, May 21, 2013
German Court Rules That Google Can Be Held Liable for Defamatory Auto-complete Search Suggestions
Anyone who has performed a Google search in the past several years will likely have noticed Google’s auto-complete function, which automatically suggests search terms as the user types. For example, when a user recently entered the terms “Brann & Isaacson,” Google suggested “Brann & Isaacson law firm.” But what happens if, instead of helpful or innocuous suggestions, Google suggests something scandalous?
Google itself has consistently maintained that it has no direct control over its results or suggestions, which are generated automatically. Because Google’s suggestion algorithm reflects the frequency with which particular search terms are entered, however, it is a natural engine for spreading and legitimizing gossip. The more a rumor is repeated, re-tweeted or re-blogged, and the more it is searched, the louder it becomes, and the more likely it is to find its way into Google’s suggested auto-complete terms. Now, a German Court has ruled that Google can be held liable if its auto-complete function suggests defamatory search results.
In Decision VI ZR 269/12 (May 14, 2013), The German Federal Court of Justice (Bundesgerichtshof), Germany’s highest court of ordinary jurisdiction, considered a case brought against Google by two plaintiffs, a company and the company’s chairman, who had sued to remove auto-complete suggestions the plaintiffs considered defamatory. Users who searched for plaintiffs’ names would see suggested search terms including “Scientology” and “Fraud.” The Court ruled that, while Google has no obligation to proof its auto-complete results in advance, it does have an obligation, once it has been put on notice that suggestions falsely imply a factual link between an individual or entity and terms that have negative connotations, to remove those terms from the suggestions and to prevent similar suggestions from appearing in the future.
The case is making headlines in Germany, partly because of its potential effect on a high profile lawsuit brought by Germany’s former first lady, who has sued Google under the same theory, alleging that Google’s suggested search terms gave credence to certain rumors about her past. At this blog, however, we will be paying particular attention to how the ruling plays out in a larger context, and whether any other jurisdictions adopt similar positions.
To date, courts in the United States have rejected attempts by private litigants to hold Google or other search engines liable for unwanted search results or suggestions. Section 230 of the Communications Decency Act provides interactive computer service providers with broad immunity against liability for third-party-created content, providing that “no provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider,” and that “no cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section.” 47 U.S.C. §230(c)(1), (e)(3). This section makes it difficult to hold a search engine liable for defamation in the United States based on search results. A woman in Wisconsin, who objects to the appearance of results and advertisements for impotency drugs in connection with her name, has filed a series of lawsuits against search engines alleging trademark violation and commercial misappropriation, rather than defamation. Her efforts have thus far proven unsuccessful. (See, most recently, Stayart v. Google, 710 F.3d 719 (7th Cir. 2013)).
Google itself has consistently maintained that it has no direct control over its results or suggestions, which are generated automatically. Because Google’s suggestion algorithm reflects the frequency with which particular search terms are entered, however, it is a natural engine for spreading and legitimizing gossip. The more a rumor is repeated, re-tweeted or re-blogged, and the more it is searched, the louder it becomes, and the more likely it is to find its way into Google’s suggested auto-complete terms. Now, a German Court has ruled that Google can be held liable if its auto-complete function suggests defamatory search results.
In Decision VI ZR 269/12 (May 14, 2013), The German Federal Court of Justice (Bundesgerichtshof), Germany’s highest court of ordinary jurisdiction, considered a case brought against Google by two plaintiffs, a company and the company’s chairman, who had sued to remove auto-complete suggestions the plaintiffs considered defamatory. Users who searched for plaintiffs’ names would see suggested search terms including “Scientology” and “Fraud.” The Court ruled that, while Google has no obligation to proof its auto-complete results in advance, it does have an obligation, once it has been put on notice that suggestions falsely imply a factual link between an individual or entity and terms that have negative connotations, to remove those terms from the suggestions and to prevent similar suggestions from appearing in the future.
The case is making headlines in Germany, partly because of its potential effect on a high profile lawsuit brought by Germany’s former first lady, who has sued Google under the same theory, alleging that Google’s suggested search terms gave credence to certain rumors about her past. At this blog, however, we will be paying particular attention to how the ruling plays out in a larger context, and whether any other jurisdictions adopt similar positions.
To date, courts in the United States have rejected attempts by private litigants to hold Google or other search engines liable for unwanted search results or suggestions. Section 230 of the Communications Decency Act provides interactive computer service providers with broad immunity against liability for third-party-created content, providing that “no provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider,” and that “no cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section.” 47 U.S.C. §230(c)(1), (e)(3). This section makes it difficult to hold a search engine liable for defamation in the United States based on search results. A woman in Wisconsin, who objects to the appearance of results and advertisements for impotency drugs in connection with her name, has filed a series of lawsuits against search engines alleging trademark violation and commercial misappropriation, rather than defamation. Her efforts have thus far proven unsuccessful. (See, most recently, Stayart v. Google, 710 F.3d 719 (7th Cir. 2013)).
Thursday, April 4, 2013
New Best Practices for Trademark Owners in an Era of Ever Increasing gTLDs
As new domains become available, trademark owners with an internet presence should consider adopting new best practices to protect their marks. By way of background, a generic top level domain (“gTLD”) is the alpha numeric string that appears after the “dot” in a web site URL. The most ubiquitous such creature is the well-known “.com.” In the 1980s, the Internet Corporation for Assigned Names and Numbers (ICANN)–the mysterious not-for-profit corporation under contract with the US Department of Commerce to see to the proper functioning of the Internet’s system of addresses–created seven gTLDs: .com, .edu, .gov, .int, .mil, .net, and .org.. In years following the creation of the original gTLDs, discussions concerning additional gTLDs led to the selection in November 2000 of seven new gTLDs for introduction: .biz, .info, .name, .pro, .aero, .coop, and .museum. In 2003, ICANN initiated a process that resulted in the introduction of six new gTLDs (.asia, .cat, .jobs, .mobi, .tel and .travel). Most recently, .xxx has been added to the mix.
At each step along the way, the introduction of new gTLDs created anxiety for trademark owners concerned that each new gTLD opened up opportunities for domain name brokers to hijack or cybersquat on second level domains consisting of valuable trademarks. For instance, acme.com might find that there are cybersquatters hosting sites at acme.org or acme.asia. In response to these concerns, at each launch of a new set of gTLDs, ICANN implemented a so-called “sunrise” period, enabling trademark owners to get first dibs on any new domain name consisting of a trademark owner’s trademark.
During the late 1990’s and early 2000’s, this led to a “best practices” regime of domain name management that consisted primarily of trademark owners registering thousands of domain names consisting of their trademarks in an effort to prevent others from doing so.
Given developments in the gTLD area in the last two years, however, this approach is no longer workable.
In late 2013, ICANN is set to launch, at last count, 1,930 additional gTLDs. The process will include an enhanced version of the sunrise period, permitting trademark owners to register their trademarks in a clearing house system. Even so, the costs associated with registering even a single trademark in each new gTLD will geometrically multiply the domain name portfolio budgets of many brands.
So what is a diligent trademark owner to do? How about nothing? With the caveat that this suggestion is likely to be viewed as heretical among my fellow IP/Internet law attorneys, it is a notion that bears some consideration. At the beginning of the Internet era, the business community lacked insight into the manner in which users would come to navigate the web. In addition, few could have anticipated the sophistication that has now been achieved by search engines in matching user queries with relevant search results.
As a result of emerging user patterns, for example, virtually no user navigates to a website by entering the URL in the location bar of a browser, even if the user knows the URL. Rather, the user enters the common name for the brand or other item in the search box of a search engine. In turn, search engines ruthlessly index web sites based on relevance criteria that make it virtually impossible for a user seeking the Acme web site, for example, to be directed anyplace else. The massive proliferation of gTLDs will only accelerate the emphasis on this relevance analysis. Accordingly, it seems likely that the increasing proliferation of gTLDs will actually make it less important, not more, for brand owners to lock up every possible variation of their marks in connection with every gTLD. If so, brand owners may derive an unexpected benefit from the additional gTLDs in the form of smaller domain name registration budgets—not larger.
At each step along the way, the introduction of new gTLDs created anxiety for trademark owners concerned that each new gTLD opened up opportunities for domain name brokers to hijack or cybersquat on second level domains consisting of valuable trademarks. For instance, acme.com might find that there are cybersquatters hosting sites at acme.org or acme.asia. In response to these concerns, at each launch of a new set of gTLDs, ICANN implemented a so-called “sunrise” period, enabling trademark owners to get first dibs on any new domain name consisting of a trademark owner’s trademark.
During the late 1990’s and early 2000’s, this led to a “best practices” regime of domain name management that consisted primarily of trademark owners registering thousands of domain names consisting of their trademarks in an effort to prevent others from doing so.
Given developments in the gTLD area in the last two years, however, this approach is no longer workable.
In late 2013, ICANN is set to launch, at last count, 1,930 additional gTLDs. The process will include an enhanced version of the sunrise period, permitting trademark owners to register their trademarks in a clearing house system. Even so, the costs associated with registering even a single trademark in each new gTLD will geometrically multiply the domain name portfolio budgets of many brands.
So what is a diligent trademark owner to do? How about nothing? With the caveat that this suggestion is likely to be viewed as heretical among my fellow IP/Internet law attorneys, it is a notion that bears some consideration. At the beginning of the Internet era, the business community lacked insight into the manner in which users would come to navigate the web. In addition, few could have anticipated the sophistication that has now been achieved by search engines in matching user queries with relevant search results.
As a result of emerging user patterns, for example, virtually no user navigates to a website by entering the URL in the location bar of a browser, even if the user knows the URL. Rather, the user enters the common name for the brand or other item in the search box of a search engine. In turn, search engines ruthlessly index web sites based on relevance criteria that make it virtually impossible for a user seeking the Acme web site, for example, to be directed anyplace else. The massive proliferation of gTLDs will only accelerate the emphasis on this relevance analysis. Accordingly, it seems likely that the increasing proliferation of gTLDs will actually make it less important, not more, for brand owners to lock up every possible variation of their marks in connection with every gTLD. If so, brand owners may derive an unexpected benefit from the additional gTLDs in the form of smaller domain name registration budgets—not larger.
Tuesday, April 27, 2010
Cybersquatting and the UDRP
Recently, I was a guest lecturer at the trademark law class at the University of Maine School of Law, invited by my colleague, Rita Heimes, director of the Maine Center for Law and Innovation. I always enjoy an opportunity to discuss trademark law in an academic setting, and it is invigorating to meet aspiring young trademark lawyers.
The topic for my talk was the Uniform Dispute Resolution Policy (“UDRP”). For those unfamiliar with this policy, it helps protect businesses against “cybersquatting.” For instance, if the owner of the Acme brand discovers that the domain name "acme.com" is registered by someone else, and is being used in bad faith, the UDRP is the most effective tool available for Acme to recover that domain name. Prior to the creation of the UDRP, the prospects for success in a domain name dispute were uncertain, and the costs were potentially exorbitant. In the mid to late 1990's, when widespread use of the Internet was first becoming prevalent, it was unclear whether existing trademark doctrine enabled the recovery of a domain name in this manner, and in any event, the only way to find out was to launch an expensive and time consuming lawsuit. This time period was typified by a “gold rush” for ownership of domain names and many famous brand owners discovered to their dismay that their best option for securing the all-important “.com” domain name was to pay millions of dollars to the prescient person who had beaten them to the punch.
The UDRP was introduced at the insistence of brand owners by the Internet Corporation for Assigned Names and Numbers (“ICANN”), the entity that governs the allocation of Internet "real estate," in order to address the gap left by preexisting law. Any person registering a domain name must consent to participation in the UDRP’s form of alternative dispute resolution, and agree to abide by the dispute resolution’s results. Typical costs to recover a domain name using this process are between $1500 and $3000, the success rate is very high (around 85%), and the process can be completed in a month or less. The UDRP has been around for more than 10 years and has been used very effectively by brand owners to recover millions of improperly registered or used domain names. Accordingly, it no longer qualifies for "cutting edge" status, although it is surprising how many brand owners remain unfamiliar with it.
In the course of preparing my remarks, however, I also had occasion to delve into some of the remaining challenges in this area, and to explore recent noteworthy developments.
For example, the increasingly widespread use of privacy or proxy shields in connection with domain name ownership threatens a number of important policy interests, and complicates the ability to police those who are habitual cybersquatters. Many persons who register domain names now do so through third parties who hold the domain names essentially in trust, but have no real control over content or other substantive behavior that occurs on the website to which the domain name resolves. Not surprisingly, this makes it difficult for those who are injured by the content appearing on the website to get access to the real party in interest. In the cybersquatting area, it can also complicate the ability to prove bad faith, since one accepted method of doing so is to provide evidence of a pattern of cybersquatting behavior by the individual respondent.
Another interesting development in the UDRP decisions also relates to the concept of bad faith. In an increasing number of cases, UDRP panels are imposing upon cybersquatters an obligation to perform some due diligence in order to determine whether registration of a particular domain name infringes the rights of a third party. If this trend gets traction in the UDRP world, it has the potential to expand significantly the scope of registrations that might be subject to a bad faith argument.
Finally, in an interesting U.S. District Court case arising out of an Anticybersquatting Consumer Protection Act (15 U.S.C. § 1125(d)) case, a judge has reaffirmed the rule that so-called "gripe sites" are protected under the First Amendment guarantee of free speech, and in most circumstances will not be subject to a claim of bad faith. Although not a UDRP case, the reasoning is likely to be followed by UDRP panels, at least in the United States. In Career Agents Network v. careeragentsnetwork.biz, the court held that a site which has no commercial purpose, but merely contains commentary and criticism, is protected. The essence of the court's holding is that there needs to be some profit motive on the part of the owner of the gripe site in order for the trademark owner to prevail. This will no doubt be disappointing to brand owners, as gripe sites can be at best an annoyance, and at worst a significant threat to brand loyalty.
The topic for my talk was the Uniform Dispute Resolution Policy (“UDRP”). For those unfamiliar with this policy, it helps protect businesses against “cybersquatting.” For instance, if the owner of the Acme brand discovers that the domain name "acme.com" is registered by someone else, and is being used in bad faith, the UDRP is the most effective tool available for Acme to recover that domain name. Prior to the creation of the UDRP, the prospects for success in a domain name dispute were uncertain, and the costs were potentially exorbitant. In the mid to late 1990's, when widespread use of the Internet was first becoming prevalent, it was unclear whether existing trademark doctrine enabled the recovery of a domain name in this manner, and in any event, the only way to find out was to launch an expensive and time consuming lawsuit. This time period was typified by a “gold rush” for ownership of domain names and many famous brand owners discovered to their dismay that their best option for securing the all-important “.com” domain name was to pay millions of dollars to the prescient person who had beaten them to the punch.
The UDRP was introduced at the insistence of brand owners by the Internet Corporation for Assigned Names and Numbers (“ICANN”), the entity that governs the allocation of Internet "real estate," in order to address the gap left by preexisting law. Any person registering a domain name must consent to participation in the UDRP’s form of alternative dispute resolution, and agree to abide by the dispute resolution’s results. Typical costs to recover a domain name using this process are between $1500 and $3000, the success rate is very high (around 85%), and the process can be completed in a month or less. The UDRP has been around for more than 10 years and has been used very effectively by brand owners to recover millions of improperly registered or used domain names. Accordingly, it no longer qualifies for "cutting edge" status, although it is surprising how many brand owners remain unfamiliar with it.
In the course of preparing my remarks, however, I also had occasion to delve into some of the remaining challenges in this area, and to explore recent noteworthy developments.
For example, the increasingly widespread use of privacy or proxy shields in connection with domain name ownership threatens a number of important policy interests, and complicates the ability to police those who are habitual cybersquatters. Many persons who register domain names now do so through third parties who hold the domain names essentially in trust, but have no real control over content or other substantive behavior that occurs on the website to which the domain name resolves. Not surprisingly, this makes it difficult for those who are injured by the content appearing on the website to get access to the real party in interest. In the cybersquatting area, it can also complicate the ability to prove bad faith, since one accepted method of doing so is to provide evidence of a pattern of cybersquatting behavior by the individual respondent.
Another interesting development in the UDRP decisions also relates to the concept of bad faith. In an increasing number of cases, UDRP panels are imposing upon cybersquatters an obligation to perform some due diligence in order to determine whether registration of a particular domain name infringes the rights of a third party. If this trend gets traction in the UDRP world, it has the potential to expand significantly the scope of registrations that might be subject to a bad faith argument.
Finally, in an interesting U.S. District Court case arising out of an Anticybersquatting Consumer Protection Act (15 U.S.C. § 1125(d)) case, a judge has reaffirmed the rule that so-called "gripe sites" are protected under the First Amendment guarantee of free speech, and in most circumstances will not be subject to a claim of bad faith. Although not a UDRP case, the reasoning is likely to be followed by UDRP panels, at least in the United States. In Career Agents Network v. careeragentsnetwork.biz, the court held that a site which has no commercial purpose, but merely contains commentary and criticism, is protected. The essence of the court's holding is that there needs to be some profit motive on the part of the owner of the gripe site in order for the trademark owner to prevail. This will no doubt be disappointing to brand owners, as gripe sites can be at best an annoyance, and at worst a significant threat to brand loyalty.
Subscribe to:
Posts (Atom)