Saving the Internet or linking limbo? CJEU clarifies legality of hyperlinking (C-160/15, GS Media v Sanoma)
By Bernd Justin Jütte
In a much awaited decision, the CJEU has ruled that linking to freely-available copyrighted content that has been uploaded without the consent of the right holder is, in principle, legal. However, it qualified that such a reference could infringe the right to communication to the public under Article 3(1) of the Information Society Directive (Directive 2001/29/EC) if certain elements were present. After AG Wathelet had provided his Opinion in early April (see post on this Blog here), the CJEU rendered its Judgement on 8 September 2016.
The facts of the case have already been summarized in the post on the AG’s Opinion on this blog and are only briefly recalled here. GS Media operates the website GeenStijl on which it had posted a link that referred users to unpublished nude pictures of a Dutch celebrity on another website. The publisher of Playboy, who held the rights to these pictures, sued for copyright infringement, arguing that GS Media had made a communication to the public of the pictures within the meaning of Article 3(1) by posting the links on its website.
The questions referred to the CJEU by the Dutch Supreme Court essentially seek to clarify if, and under which circumstances, a hyperlink to content protected by copyright can constitute a communication to the public within the meaning of Article 3(1) of Directive 2001/29/EC. The right to communication to the public is one of the exclusive rights granted under the provisions of the Information Society Directive. It gives authors “the right to authorize or prohibit any communication to the public of their works by wire or wireless means, including the making available to the public of their works in such a way that members of the public may access them from a place and at a time individually chosen by them.“ Article 3 has generated by far the most preliminary references of all of the exclusive rights contained in the Directive. Two of these references dealt specifically with linking and framing on the Internet. In distinction to earlier case-law, this preliminary reference relates to content that is freely available on the Internet, but which has not been made available with the consent of the right holder. The answer to these questions closes a systematic gap that had been left after the judgments in Svensson and BestWater, which both related to content which had been made available with the consent of the right holder.
The AG’s Opinion
The AG had argued that links such as those in the main proceedings did not amount to a communication to the public. In a long line of case-law the Court of Justice has established that Article 3(1) of Directive 2001/29/EC is composed of two elements: an ‘act of communication’ which is directed to a ‘new public’. First, the AG broke with the established interpretation of the CJEU, opining that setting a hyperlink does not constitute an act of communication because links do not make works available, but merely help to access them more easily. Second, he suggested a much simpler test to decide whether an act of communication is directed to a new public: This test would require an examination of whether a link is indispensable (and not merely facilitating) for members of the public to access works protected by copyright. He further argued that no other criteria should be taken into consideration.
The Court did not follow the AG’s Opinion but clung to the more established interpretation of Article 3(1). Before going into the substance of the analysis, the Court recalls the nature and purpose of the right of communication to the public. It underlines the preventive nature of the exclusive right and stresses that its precise scope must be defined by interpretation of the text of Directive 2001/29/EC. In doing so, the Court must guarantee a high level of protection for right holders (Recitals 9 &10), while maintaining a fair balance between the interests of right holders and users of protected works (Recitals 3 & 31). Based on the previous case-law of the Court (incl. FAPL/Murphy, to SGAE v Rafael Hoteles and SCF v Marco del Corso) an act of communication to the public consists of two elements (para. 32): (1) an ’act of communication’ (2) to a ‘new public’. The existence of an act of communication to the public has to be assessed individually on the basis of a set of complementary and interdependent criteria. In this assessment a court has to consider the role of the user and whether his intervention was made knowingly and whether it was necessary to give access to the content; a court has to further ask whether the act of communication to a ‘public’ is made to an indeterminate number of individuals where that number has a certain magnitude, and whether this communication uses technical means different from those of the original act of communication; and finally, whether the act of communication to the public serves to generate profits for the user. (see paras. 33-9)
Back to square one: linking is communication
The CJEU argues, though not expressly, that setting a hyperlink constitutes an act of communication. This confirms its prior jurisprudence on the interpretation of Article 3(1) (see Svensson and BestWater) but rejects the argument made by the AG that setting hyperlinks does not constitute an act of communication as the first element the provision. It qualifies this statement, stating that Svensson and BestWater concerned links to (freely available) works that were made available with the consent of the respective right holders. (para. 41) Therefore, these communications were not directed towards a new public. However, according to the Court, this does not automatically mean that links to content which is freely available on the Internet, but has been made available without the consent of the right holder, do not constitute a communication to a new public. As a result, whether content is already available on the Internet without any access restriction does not clearly answer the question whether a communication to a new public takes place. This reasoning seems to be in contradiction to the rulings in Svensson and BestWater; since in all three cases (incl. GS Media) the linked content was already freely available without access restrictions on the Internet at the time the links were set. But the Court argued that both earlier judgments, which established the legality of linking to and framing of content which has been made available with the consent of the right holder, merely underline the importance of the right holder’s consent for the legality of such references. (paras. 41-3). Accordingly, links to freely available content that has been posted without the consent of the right holder could constitute an (infringing) act of communication to the public if they have been posted to content which has been made available without the consent of the right holder.
Freedom of expression saves the day (viz. Internet)
Such an interpretation of Article 3(1), which potentially makes linking to infringing content an act that infringes the right of communication to the public, would have created (too) much legal uncertainty. Therefore, the CJEU qualifies this interpretation with reference to the right to freedom of expression. The Court underlines the importance of the Internet for the exercise of this fundamental right. And therefore, it argues, qualifying all links to content that has been posted without the permission of the right holder as a communication to the public would have a chilling effect on the exercise of the right to freedom of expression on the Internet. In order to conduct an individual assessment of a potential act of communication to the public, further criteria must be considered. (paras. 44-6)
Profit and knowledge
Users who post links to ‘illegal’ content that do so without an economic motivation do not make an act of communication to the public if they cannot reasonably know that the content has been uploaded without permission from the right holder. However, upon gaining knowledge, e.g. by notification by the right holder, such a link can constitute a communication to the public. (paras. 47-9) An act of communication to the public also takes place when a user sets a link to content that is protected by a paywall or other technical restrictions, and is therefore not freely available. (para. 50) This means in turn that users who post links as part of their economic activities are presumed to have ascertained whether the linked content has been made available with the consent of the right holder. (para. 51)
Accordingly, an operator of a website who posts links to unpublished photographs as part of a commercial activity makes an act of communication to the public that requires authorization by the right holder.
With its ruling in GS Media the CJEU closes a huge gap by answering the question when and under which circumstances linking to infringing content constitutes an act of communication to the public within the meaning of Article 3(1) of the Information Society Directive. The judgment, unfortunately (in this author’s opinion), did not follow the brave and systematically more elegant suggestions made by AG Wathelet in his Opinion, and thereby missed an opportunity to alter the CJEU’s rulings in Svensson and BestWater. Not only did the Court reject the AG’s suggestion not to qualify hyperlinks as an act of communication, but it also discarded his suggestion of a ‘meta-criterion’ of ‘indispensability’ to decide whether a communication is directed towards a new public. Instead, the Court adds to the ever-growing catalogue of elements that are constitutive (or not) of an ‘act of communication to the public’, their ‘interdependence’ further complicates any analysis:
These constitutive criteria now include:
- an act of communication
- to a new public
- knowledge of the potential ‘illegal’ nature of the original source
- profit-making nature of the hyperlinking website
In particular the latter two elements might prove problematic when determining whether an act of communication to the public actually occurred. Whereas there might (in the future) be a bright-line rule for the commercial, or profit-making nature of a given website (e.g. income from advertisement, even if marginal, main economic activities, etc.), the criterion that pertains to the knowledge (of the lack of consent of the right holder) of the non-commercial user might prove much more problematic. It should be further clarified when non-commercial users who link to infringing content are liable without receiving notice for the links they post, or whether there is a general assumption that all hyperlinks set by private users are non-infringing. The Court expresses itself rather vaguely by stating that a user might make an act of communication to the public if he “knew or ought to have known that the hyperlink he posted provides access to a work illegally placed on the internet“. (para. 49) In contrast to the expectation that commercial users “[carry] out the necessary checks to ensure that the work concerned is not illegally published on the website to which those hyperlinks lead” (para. 51), private users do not incur such an obligation. A knowledge as to the infringing nature of certain content would then have to be established based solely on the appearance of the linked website.
Given the vital importance and ‘indispensability’ of hyperlinks for the functioning of the Internet, and therefore, as the Court correctly noted, the exercise of a number of fundamental rights, the uncertainty created by the interpretation of Article 3(1) might indeed have chilling effects on the exercise of some of these rights. Users without any legal knowledge and, at best, an anecdotal knowledge of copyright law, will have difficulties to determine which content is legal, and can be linked to, and which is not, and can therefore not be linked to. To avoid potential liability users should rather not post links to questionable content. Often such checks will be impractical or simply infeasible to conduct in a short amount of time, which might hamper the dynamic and fast dissemination of information.
Furthermore, users will also have difficulties to determine whether their activity is of a profit-making nature, which would require them to ascertain the legality of the content they intend to link to. In the absence of precise criteria for ‘profit-making’ a number of scenarios can be conceived where this criterion could lead to uncertainty. For example, the activity of a blogger whose website displays advertising in order to finance the registration of his website could be considered as profit-making, as the advertising generates income (which is immediately consumed by the running costs of the website. Or even if this blogger earned a small amount of money with this advertising, would this require him to ascertain the legality of his sources?
There is still too much uncertainty created by the multitude of criteria established by the Court. However, as it seems, these criteria will shape the legal environment for hyperlinks (and therefore the Internet) in the near future.
Moreover, as also noted by The IPKat’s Eleonora Rosati, the judgment de facto establishes a take-down obligation for (even non-commercial) users. According to the Judgement, the moment a user is informed that a hyperlink directs users to works that have been published without the authorization of the right holder, a non-infringing hyperlink morphs into an infringing (non-authorized) act of communication to the public. To escape liability, a user will have to take down the link to the infringing content. Upon failure to do so users would then incur direct liability. Such a notice-and-take-down system, thus far, only exists for passive hosting providers under Article 14 E-Commerce Directive (see Recital 42), which establishes secondary liability for intermediary service providers if they fail to remove infringing content once they have been notified.
The fun (for IP lawyers) is far from over as at least two relevant cases are still pending before to the Luxembourg Court (Stichting Brein I and Stichting Brein II) on the interpretation of Article 3 of the Information Society Directive. The first deals with hyperlinks to infringing content that are provided with a media player, the second with websites that do not provide infringing content but provide a metadata which enables users to download illegal content from other users.