We’re happy to bring you another guest post by our Fellowship applicant Prarthana Patnaik. Prarthana is a 4th year student at National Law University, Odisha. This is her third submission for the Fellowship. The previous submissions can be viewed here and here.
Google Images’ ‘View Image’ Button and Copyright Infringement
Google’s Image Search service allows users to view thumbnail images (low-resolution reproduction of images available on other websites) in response to the specific search query entered by the users. It uses a “spider” software which ‘crawls’ the web and creates an index of thumbnails which are then displayed through an index-searching software. The ‘View Image’ button that was part of the Image Search service until recently, allowed the users to directly access and download the full-resolution version of the thumbnail, bypassing the hosting website. This feature of its service attracted a lot of controversy since it facilitated users to steal copyrighted images and also reduced viewership and traffic of the hosting websites.
In the wake of this, Google recently decided to remove the ‘View Image’ button. While this has been perceived as a victory for copyright holders (photographers, publishers etc.), netizens, who say that this was one of the most useful feature of the Image Search service offered by Google, are complaining against it. So, why did Google remove the ‘View Image’ button? Did it apprehend accusations of copyright infringement for providing this feature?
In the past, the Image Search services offered by Google and other visual search engines have been the subject of litigation in USA. In Kelly v. Arriba Soft Corp (‘Kelly’), in 2003, the US Court of Appeals for the Ninth Circuit examined whether Arriba, an internet search engine, infringed copyright in Kelly’s photographs by ‘crawling’ the web looking for images, creating thumbnails of those images (including Kelly’s photographs) and then displaying them in response to search queries. On clicking a thumbnail, users were directed to the original hosting website where they could view and download the full-sized image. On applying the four-factor fair use test, the Court held that Arriba’s use of Kelly’s images as thumbnails constituted fair use. It noted that although the use was commercial, it was highly ‘transformative’ in nature as the photographer’s use of his images was for aesthetic and artistic purposes whereas Arriba’s use of those images was for a functional purpose of creating a catalogue and improving access to artistic works. With respect to market harm, it observed that Arriba’s thumbnails did not harm the market/value for Kelly’s full-sized images because thumbnails, which lose their clarity when enlarged due to their low resolution, did not act as a substitute for full-sized images.
In Perfect 10 v. Amazon (Perfect 10), in 2007, the same Court examined whether Google’s creation and display of thumbnails of Perfect 10’s erotic images and linking to third-party sites where infringing full-size version of these images were displayed amounted to direct and secondary/contributory copyright infringement, respectively and if it did, whether it fell within the ambit of fair use. Applying the ‘Server Test’, the Court noted that since Google’s computers stored thumbnails of Perfect 10’s images and communicated their copies to the users, Perfect 10 had succeeded in prima facie proving that Google’s thumbnails directly infringed its display right. However, applying the same test in respect of full-sized images, it held that since the images were neither stored on Google’s servers nor were their copies (but only HTML addresses of them) communicated to the users’ computers, Google neither ‘displayed’ nor ‘distributed’ the images to its users but only directed them to websites which stored the images.
The Court then went on to apply the fair use doctrine to assess whether Perfect 10’s use of thumbnails constituted fair use and upon a similar four-factor analysis as in Kelly, concluded that it did.
In examining Google’s secondary liability for linking to infringing full-size images displayed on third-party websites, the Court first found that there had been direct infringement by these parties since they had made unauthorized use of Perfect 10’s images. It then examined Google’s contributory liability (i.e. whether Google, with knowledge of infringement, intentionally induced or encouraged direct infringement) and vicarious liability (i.e. whether Google ‘controlled’ the infringing party and derived “direct financial benefit” from the direct infringement). Although the Court remanded the claim of contributory infringement to the District Court, it concluded that Google could not be held vicariously liable because Perfect 10 had not proven Google’s right to stop or limit direct infringement.
Google Image Search’s ‘View Image’ Function: Copyright Infringement or Fair Use?
The decisions discussed above make it clear that Google’s creation and display of thumbnails does not constitute copyright infringement in the US since it falls under the fair use exception. However, its liability for making full-size, high resolution versions of copyrighted images directly available for viewing and downloading, through its ‘View Image’ feature is not clear.
Before we delve into the issue of whether the ‘View Image’ feature could constitute infringement or not, it is necessary to note a few differences between its functions and the functions of the image search service provided by Arriba (later renamed Ditto) in Kelly and Google in Perfect 10. In the latter cases, clicking on a thumbnail enabled users to access and download full-size images only by visiting the hosting website. However, the ‘View Image’ function acted as a direct link between the thumbnails to the full-size image stored on the hosting website’s server, allowing users to thereby bypass the hosting website. In effect, when one clicked on the ‘View Image’ button, one was not led to Google or the hosting website, but directly to the image stored in the hosting website’s server, which was enabled by the HTML instructions provided by Google.
A. United States
If we apply the test of direct infringement as applied in the Perfect 10 case to the ‘View Image’ feature, we arrive at the same conclusion that there is no infringement, since it is clear that by providing access to the full-size image, Google neither ‘displayed’ (as it did not store a copy of the image in its servers) nor ‘distributed’ (as it did not ‘actually communicate’ the images to user’s servers but merely gave the address of the image to user’s browser and the user’s browser interacted with the web browser which stores the copyrighted image) the images hosted by others. However, it has to be noted that the applicability of the Server Test in respect of display rights has been recently questioned by the New York District Court in Goldman v. Breibart on the basis of the meaning of ‘display’ under the US Copyright Act. In this case, the court concluded that ‘display’ did not necessarily require a ‘copy’ of the image to be stored in a tangible medium of expression and it can be covered by any ‘device or process’ alone.
In applying the tests of secondary infringement to the Google ‘View Image’ feature, we see that Google assisted websites to enable users to access their images (infringing or otherwise) with its ‘View Image’ function. In cases where the plaintiff can prove that Google had knowledge of infringing images on its search engine and that it could take simple measures to prevent damage to plaintiff’s copyrighted works and failed to take such steps, a claim of contributory infringement would lie. Also (though this was discussed in the context of vicarious liability and then held as a claim of contributory liability), Google’s practical ability to police such infringing conduct needs to be looked into. In Perfect 10, the Court held that Google lacks such ability since it did not have image recognition technology. However, such is not the case today, as can be seen by Google’s reverse image search technology. Similarly, on applying the test of vicarious liability, such liability will only arise when Google’s agreement with the third party website is such that it has the legal right or ability to prevent it from committing acts of infringement and that it derives “direct financial benefits” from such use of images by third parties. Hence, secondary liability of Google will differ from case to case.
Assuming that linking to full-size images prima facie amounts to copyright infringement, in my opinion, it does not fall under the fair use exception In examining the first factor, Google’s use of the full-size images would not be construed as transformative since the full size images were mere reproductions of the original images and completely superseded the original work. Also, the purpose of accessibility to images was fulfilled by the thumbnails alone and there was no need to make full sized images available for viewing and downloading to fulfil such purpose. Hence, since the ‘View Image’ button does not “add something new, with a further purpose or different character”, the first factor would weigh against Google. The second factor would weigh slightly in favour of the copyright owner as the original images are creative, yet published works. The third factor would also weigh in favour of the owner since, when one looks into the purpose of copying, one finds there is no reasonable relation between the quantity of the material used and the purpose behind doing so i.e., the purpose of furthering accessibility was already served by thumbnails. The fourth factor would also weigh against Google since the full-size images would harm the market for the owner’s images, by directly substituting them. Also, the ‘View Image’ function allowed the user to bypass the hosting website and the same led to decreased traffic and viewership in the owner’s website (such harm is clear from Google’s settlement with Getty Images). On viewing the factors as a whole, Google’s ‘View Image’ function would not fall under the fair use exception, in my opinion.
The images (both the thumbnails and the full-sized images) would be considered as “artistic works” under Section 2 (c) of the Indian Copyright Act, 1957. Section 51 deals with infringement and the Google’s function could either fall under Section 51(a)(i) (direct infringement) and/or Section 51(a)(ii) (secondary infringement).
As per Section 51(a)(i), direct infringement occurs when, one does any act, which the copyright owner has the exclusive right to do. The exclusive rights of an owner of an artistic work are listed under Section 14(c) and out of them, the rights that could be alleged to have been infringed by Google are “right to reproduce the work in any material form”, “right to issue copies of the work to the public” and the “right to communicate the work to the public”. Google’s View Image function cannot be said to violate the reproduction right since the images are not stored by Google in any medium by electronic means i.e., it is not stored in Google’s servers and remains in the hosting website’s server. It cannot be said to infringe the right to issue copies to the public since Google does not issue ‘copies’ of the images but merely provides a link to the images. It may infringe the communication right, however. Section 2(ff) of the Act states that “communication to the public” means “making any work available for being seen or heard or otherwise enjoyed by the public directly or by any means of display or diffusion other than by issuing copies of such work regardless of whether any member of the public actually sees, hears or otherwise enjoys the work so made available.” This right has mostly been invoked in the context of broadcast of cinematographic works and not in respect of artistic works yet. Also, focus has been laid on the meaning of ‘diffusion’, instead of that of ‘display’. If ‘display’ is taken to be similar as ‘display’ in the US copyright law and includes making of ‘copies’ which are “fixed in a tangible medium of expression” only, then Google’s ‘View Image’ feature would not amount to infringement of the owner’s communication right. However, if ‘display’ is interpreted to include electronic modes of communication, then it could constitute infringement.
According to Section 51(a)(ii), when one permits for profit any place to be used for communication of copyrighted work to the public, with the knowledge/reasonable belief that such communication constitutes infringement, one commits secondary infringement. Further, it has been held by the Delhi High Court in the MySpace case that in order to be liable, the intermediary must have “actual knowledge” of “specific infringing material”. Therefore, in each case it would have to be seen whether Google had actual knowledge of the availability of specific infringing material through service and whether it reaped any profit from it depending on its agreement with third parties. If all of this could be proven, a claim of secondary infringement could lie against Google for its Image Search service including the ‘View Image’ feature.
Assuming that there is a prima facie case of infringement in India, one would then have to examine whether Google’s ‘View Image’ function can fall under the Indian version of the fair use doctrine as codified under Section 52 of the Copyright Act, 1957. The structure of the Indian provision differs from its American counterpart. Unlike Section 107 of the US Copyright Act that merely lays down the four-factor test, Section 52 lists out an exhaustive list of acts that do not constitute copyright infringement. Since Google’s Image Search service does not fall within the ambit of any of the acts excluded under Section 52, Google’s use of full-sized version (‘View Image’ feature) as well as thumbnails of copyrighted images could constitute copyright infringement in India if it amounts to ‘display’ of those images.