Search engine liability for third party infringement: a keenly awaited ruling
The Argentine Supreme Court has held that search engine liability arising from third parties’ infringing content must not be determined under a strict liability regime but should be decided on the principles of fault-based liability.
The Argentine Supreme Court stated that search engines could only be found liable for third parties’ infringing content if they failed to take corrective steps upon having actual and effective knowledge of the unlawful content. The Supreme Court also held that, since search engines do not have monitoring obligations to control websites' content, a strict liability regime must not be imposed.
In connection with the type of notice that must be served to search engines for them to have ‘actual and effective knowledge’ of third parties’ infringing content, the Supreme Court distinguished two different scenarios. In cases where the infringing nature of the content is notorious and flagrant, a warning notice must be sent by the claimant to the search engine in order to determine the ‘actual and effective knowledge’ of the infringement. However, when the infringement is questionable, doubtful or requires further analysis, a warning letter will not be enough and a notification order issued by a judicial or competent administrative authority is required to establish the ‘actual and effective knowledge’ of the search engine.
With regard to the use of ‘thumbnails’, the Supreme Court stated that this practice is not different from including links to third party websites listed on the natural search results (i.e. those results that are returned based on the natural indexing of the web site, as opposed to those that are returned based on paid advertising) in so far as search engines act in both cases as mere intermediaries. Search engines will not therefore be liable unless they have ‘actual and effective knowledge’ of the infringing content.
The minority opinion (Justices Ricardo Lorenzetti and Juan Carlos Maqueda) held that section 31 of Law 11.723 is applicable and that search engines therefore need to obtain authorization from third parties to use their images unless the consent is not needed under section 31 (no consent being necessary when the use of protected attributes appears in a publication related to scientific, educational or cultural purposes, nor is it necessary for facts or events which are of public interest or have taken place in public).
No specific rules governing search engine liability have been passed in Argentina. Due to the lack of a safe harbour system (such as that established in the European Directive on Electronic Commerce, section 230 of the Communication Decency Act and/or section 512 of the Digital Millennium Copyright Act), search engine liability relating to infringements of personal or intellectual property rights caused by third party content hinges on the general tort rules stated in the Argentine Civil Code.
In recent years, courts have had to deal with search engine liability in many cases (so-called ‘celebrity cases’) where famous people discovered that, when typing their names on Google or Yahoo!, search results showed links to pages with sexual, erotic and pornographic content and/or which used their images on the image search feature without authorization (‘thumbnails’).
In certain cases, courts applied the rule of fault-based liability and stated that search engines could not be liable if they were not previously notified of the infringing content on third party websites and had not been asked to remove the link to it. Liability would arise if search engines failed to react diligently upon the ‘effective knowledge’ of the unlawful content.
In other cases, courts applied the rule of strict liability considering that the search engine software used to list organic search results and search engine main activity are risky per se. Search engines must therefore be liable for the resulting damage independently of whether they have been notified of the infringing content. This position directly imposed upon search engines the unreasonable and impracticable obligation to check and monitor all content on the internet prior to the effective knowledge of the infringement.
Before rendering a decision in María Belen Rodriguez the Supreme Court held a hearing with the participation of an amicus curiae on 21 and 29 May 2014. The Supreme Court's ruling in this case seems to close an important discussion held in Argentina for many years in connection with the standard applicable to search engine liability arising from third parties infringing content.
The model Maria Belen Rodriguez filed a complaint against Google and Yahoo! and asked the court to order the defendants to pay damages plus interest, remove all search results associating her name with websites of sexual and pornographic content or nature and to remove all thumbnails depicting her name from the search results feature.
The court of first instance ruled in favour of the plaintiff and, based on the fault liability standard, ordered Google and Yahoo! to pay damages of ARS 100,000 and AR$ 20,000 respectively, plus interest and litigation costs. The court of first instance also ordered the take-down of all search results linking the plaintiff's name to sexual, erotic and/or pornographic website and thumbnails associating the plaintiff with these websites of a sexual, erotic and/or pornographic nature.
On appeal, Division A of the Civil Court of Appeals upheld the application of the fault-based liability standard and ruled that the use of the plaintiff's photographic image as a thumbnail was unlawful. It also established that search engine activity was protected by the constitutional right to inform and freedom of speech and released the defendants from all liability for third parties infringing content because the plaintiff had not proved that the defendants had taken corrective steps after having had actual and effective knowledge of the infringement. Moreover, the Civil Court of Appeal stressed that there was no evidence of the plaintiff sending a warning notice to the defendants, of notifying the infringing websites, and that the plaintiff opted directly to file for injunctions against Google and Yahoo! Damages imposed on Google for the unlawful use of the thumbnails were reduced to AR$ 50,000.
Both parties filed an appeal with the Supreme Court.
This ruling is extremely important because the Supreme Court established its position on the standard applicable to search engine liability, adopting the negligence standard and disregarding the strict liability regime. It also means that the court definitively established that search engines have no proactive monitoring obligations to check the content of all website on the internet.
The Supreme Court emphasized the benefits that search engines' main activity gave to society, adding that this activity is protected and guaranteed by the constitutional right to inform and to be informed and the right of freedom of speech.
The Supreme Court adopted a notice-and-take-down system with particular characteristics. The country's highest court established that search engines could only be held liable for damages caused by third parties infringing content if they have ‘actual and effective knowledge’ of the infringement and do not take any corrective steps to put end to it.
Then one question arise: When should be understood that search engines have ‘actual and effective knowledge’? The Supreme Court distinguished two cases. For those cases where the infringing nature of the content is blatant, a mere warning notice sent by the claimant is enough to consider that search engines have “actual and effective knowledge” of the infringement. By contrast, when the infringing content is doubtful, debatable or requires further analysis and investigation a warning notice sent by the claimant would not be sufficient and a notification order issued by a competent judicial or administrative authority is required.
The Supreme Court mentioned examples of cases where the infringing content is notorious and flagrant, such aschild pornography, information that facilitates the commission of crimes or puts lives or safety of one or many people at risk, statements in support of genocides, racism or any other type of discrimination, information of ongoing judicial investigations that must remain secret and classified or that constitutes libellous statements, photo-montages of clear fake images or that, in any way, constitutes a clear violation of privacy, showing photographs or images of acts that due to their intrinsic nature belong undoubtedly to the privates sphere of a person although not necessarily sexual acts.
Some questions arise. Is the list of cases mentioned by the Supreme Court exhaustive? Which government authority could be considered a competent administrative authority? These and other questions have not been answered by the Supreme Court.
We may validly affirm that, considering freedom of speech and the constitutional right to inform is involved and that search engines could not be reasonably be placed in the position of courts to determine if certain content infringes third parties rights or not, the general rule should be a notification order by a competent judicial or administrative authority and, exceptionally, a private communication by a claimant or any third party.
This important ruling has been awaited for quite a long time. Before this ruling, the standard of liability applicable to search engines for third parties infringing content in Argentina was unclear. Some cases held responsibility under a strict liability regime (see Marco Rizzo Jurado's article ‘Search engine liability arising from third parties infringing content: a path to strict liability?’ Journal of Intellectual Property Law & Practice, (2014) 9 (9): 718–720, first published online 14 July 2014) and others applied the rule of fault-based liability. With this Supreme Court precedent the strict liability regime seems to be finally disregarded. Nevertheless, there are several questions that may arise from this case. For instance: is this ruling applicable only to search engines or could it be extended to blogs, social networks or other intermediaries? Does this precedent only apply to personal right infringements or could it be extended to trade mark, copyright, trade secrets or other intellectual property right infringement? Are data protection regulations involved in this type of case and how may this case be analysed under the right to be forgotten? These and other relevant aspects remain uncertain and are likely to be treated by courts in the near future.