Search engines, which are large databases put together, acquire their data by seeking information from website operators, ISPs or their domain name registrars . Image search engines in particular consist in a database of images that are copied into thumbnail form for indexing and subsequent searching . One of the best examples of image search engines is without no doubt, the giant Google Images. Indeed, Google Images is the first stop for many of us looking for the perfect image on our blog or presentation slides. Image search engines, such as Google, have opened up a huge world of images to everybody.
As a matter of fact, image search engines can benefit from database protection provided that all the conditions are fulfilled . However, of greater controversy and importance is whether the activities of search engines can amount to copyright infringements.
- A view from the United States
Several US cases addressed the issue of copyright infringement for image search engines. It is commonly recognized that the two landmark cases are Kelly v Arriba Soft and Perfect 10 v Google .
a) Kelly v Arriba Soft
In this case , the plaintiff, Kelly, is a professional photographer who publishes his work in books and on the Internet. The defendant, Arriba (now called ditto.com or Ditto) is a visual search engine dedicated to images and provides therefore a list of reduced thumbnail pictures according to the search query.
In 1993, around 35 Kelly’s images were put in the defendant’s database and made available to the public in thumbnail form. Following the objection of the plaintiff, Arriba removed the images from its database but some reappeared due to technical problems. In the meantime, Kelly brought the case in court for copyright infringement and infringement of the US Digital Millenium Copyright Act (DMCA) by removing or altering the copyright management information associated with the plaintiff’s images.
The US District Court held that the use of copyrighted images by search engines results prima facie in copyright infringement, but is justified under the fair use exception. The Court based its decision on the ‘transformative’ nature and purposes of Arriba’s use. In fact, while Kelly’s work has been created for artistic and illustrative purpose, Arriba’s work has been created for functional purpose: to enable people to find images on the Internet. This ‘transformative’ character was clearly the decisive factor that leads the Court to rule in favour of fair use.
b) Perfect 10 v Google
Perfect 10 v Google was a US court decision between Perfect 10, the plaintiff, that is an adult men’s magazine and Google, the defendant. In 2001, Perfect 10 complained Google to cease in-line linking to full size images and creating and distributing thumbnails of its images in its Google Image Search Service, and, as a consequence, filed in 2004 a legal action against Google for copyright infringement.
The District Court opined that the use of frame and in-line linking did not amount to copyright infringement. However, the court reasoned differently with regard to the creation of thumbnails by ruling that Google was likely to be held liable of copyright infringement due to the creation and the display of these thumbnails. Here, the court did not follow the reasoning of the above-mentioned Kelly (which provided that the creation and use of thumbnails was justified under fair use), given that Perfect 10 was also present in the market of reduced-size images for download to mobile phones, and Google’s thumbnails were likely to affect the market for those reduced-size images .
In appeal, the Ninth Circuit Court of Appeals reversed the District court judgement, and, applying Kelly, held that Google’s use was fair use. The Court of Appeals essentially based its ruling on the fact that, in line with Kelly, Google’s use was significantly transformative with a different purpose which provided a considerable help for the public.
- A view from the Continental Europe
In Germany, the Hamburg regional court addressed the same issue in 2003: a case involving the claim of copyright infringement from a photographer whose images had been published by the German version of Google. It was quite simple: here, the court held that Google was liable of copyright infringement. No fair use or other defences were recognized. In 2008, the same German court addressed a similar case where Google distributed thumbnails of the images belonging to an artist and a photographer, and took the same decision.
However, few years later, a veritable turnaround occurred in Germany. Indeed, the Court reversed the above-mentioned practice and ruled in favour of Google by considering that Google’s use, though creating and indexing thumbnails of copyrighted images in its image search service, was fair use.
More surprisingly, French courts followed the German position and ruled in favour of Google . Along with the fair use defence, French courts grounded that Google could not be held liable of copyright infringement due to freedom of information.
The mere justification of freedom of information, without any development is, however, disappointing. In fact, if freedom of information can justify Google’s use, why could it not justify all magazine publishers who wish to rely on copyrighted images? Indeed, accepting that freedom of information constitutes per se a fair defence to copyright infringement not only questions the raison d’être of the whole copyright system but also opens the door to misapplications
- Does Google have an implied license?
An interesting aspect of cases involving Google for copyright infringement (cases involving Google Images, Google News, Google Book Search) is that Google has often attempted to argue that it benefits from an implied licence. Google argued therefore that “by making the articles available to the public on the Internet, the rights holders had explicitly or implicitly authorised the copying and referencing of their sites. In particular, the publishers could have used meta tags and robot.txt files to prevent Googlebots automatically copying and indixing their sites”.
In the worldwide jurisdictions, there has been consideration of this matter in a number of cases, including the US Blake A Field v Google Inc , where the Court surprisingly accepted such a defence. Similar cases from other jurisdictions, however, including the well-known Belgian case Copiepress v Google , rejected the view that the right holder who displays his copyrighted images on the Internet must use technical means to prevent automated searching, otherwise he will be considered to have authorised the operations of search engine operators.
- Does Google have a hidden impunity?
In the light of the above, it appears that the use of thumbnails images by search engines could in exceptional circumstances be considered as direct copyright infringement. Surely, copyright infringement occurs when web crawlers or comparable technologies to consult certain webpages in order to extract information to be reproduced/re-presented for commercial purposes.
Whilst Courts held that generally images search engines do not directly violate copyright, it is still arguable as to whether their activities may be considered as an authorisation of copying. In order to escape from any possible lawsuit for secondary liability, and under pressure from the creative industries, Google implemented a filtering option that lets the user know when he can use, share, or modify an image he finds via Google. While Google should be applauded rather than criticized for doing that, the impunity offered to Google, and the other search engines, is still not justified.
