Searching for a Coherent Copyright Regime in the Area of Search Engines

IPRinfo 5/2010)

Cédric Manara
Associate professor of law at EDHEC Business School, France (LegalEDHEC Research Centre)

Search engines create cached copies of search results. By doing so, do they infringe copyright law?

Did you surf on the internet today? If your answer is yes, then you probably visited if you are Chinese, if you are Japanese, or if you are North American and one of the European versions of Google if you live somewhere between Iceland and Russia.

The last time a company dared calculate the number of requests made on search tools in a month, it came up with the incredible number of 113 billion searches conducted worldwide in July 2009 (ComScore, August 2009).

Critical Crossroads of Information
Search engines are critical to our digital and offline lives. They are at the crossroads of information channels. To help us accessing what we would like to find, they use robots that never stop crawling the web to retrieve data.

Each search tool builds an index from what has been collected. Some engines also create a ’cache copy’ of the web content, from which users may read a page that is currently unavailable.

When users look for images, results are typically returned using the ’inline linking’ technique, by which search engines display thumbnails of pictures stored on third party servers. Tech-savvy web publishers can include hidden codes in their page to prevent robots from crawling or indexing their content. They rarely do so, however.

By design, search tools are based upon the use of content that is not theirs. Part of this content is subject to copyright. When search tools create copies, or return results in the form of texts or images, do they infringe copyright law?
The answer is different from country to country. This leads to undesirable legal uncertainty, an issue that should be resolved, as a secure business model for search engines means the guarantee for all us to use them for free.

US: Legal Benediction of Search Engines
In 2008, Google said its systems that process links had found one trillion unique URLs on the web. Behind them there are millions of people publishing content. Not all of them agree to the use of their content by search engines. Among them was Blake Field , who complained before a US court that Google was engaged into unauthorized copying when it created a cached version of the website where he posted poems.

He also alleged that the fact of displaying cached copies as a result to users was unauthorized distribution. The court not only ruled that Google’s disputed acts were covered by the fair use exception, but also that the person who publishes online accepts to be included in search tools index, except if notice is given of a refusal.

In other cases before US courts (Kelly v. Arriba, and Perfect 10 v. Amazon & Google), it was similarly held that the way search engines function is not contrary to the Copyright Act. Also, the so-called ’safe harbor principles’ could apply, but this issue will not be discussed here.

In countries such as Australia, Canada, India or Israel, whose legislation is more or less inspired by the American copyright regime, search engines activity may also be deemed lawful. But this is not necessarily the case in European states.

Europe: Judicial Jeopardy for Search Engines
The s.c. Infosoc Directive (EC Directive 2001/29 on copyright in the information society) provides for exceptions or limitations to the intellectual property rights of those whose works are being used, even on the web. Some temporary acts of reproduction ”which are transient or incidental and an integral and essential part of a technological process” cannot be objected to by reproduction rights holders.

This provision was drafted to protect internet service providers, but was not designed to take into account search tools activities, and cannot convincingly be seen as applying to them.

Search tools do not display excerpts of websites, resized pictures, or full titles, in a way that is fully compliant with the EU copyright regime. This opened the door to several lawsuits, mainly against the biggest search engine company Google in Belgium, France, Germany and Spain.

In Italy, there are pending (2010) proceedings against Google before the national competition authority, as a professional group of newspaper publishers complains of some of the search tool practices. This case does not directly deal with copyright and thus it will not be addressed here.

Paradoxical Ruling by a Spanish Court
In Belgium and France, it was ruled that storing content in the cache memory is illegal reproduction, and/or that displaying third parties’ content in results pages infringes their rights.

French case law is, nevertheless, divided: one court found that the applicable law should be the US Copyright Act and that the place where the copy was made should prevail, not where it was brought available. It then ruled that Google’s use of photos was fair.

A Spanish court also referred to fair use but without any reference to the American regime! After having noted that it cannot be denied that Google’s activity violates the intellectual property rights of the complainant, an artist who was unhappy to see his works in the search tools results, the Barcelona jurisdiction nevertheless ruled in favor of the American company.

This paradoxical ruling highlights the tension that exists between the copyright rules and the operation of some of the most popular services on the web.

Online Publishing as Implied License
A parallel can be drawn with a similar case that took place in Germany and went up to the Supreme Court (BGH). After lower courts had had different opinions on the lawfulness of Google Images Search, the BGH eventually ruled that the search tool could understand the complainant’s actions as consent to the display of her graphic works in search results.

Perplexity is the first reaction to this rapid survey of European national case law. Courts either apply the strict rules of copyright, which results in sanctioning search engines, or tend to be more ’creative’ to issue a decision in favor of search engines.

Among them, the German approach is the most interesting. It finds that publishing online equates to granting an implicit license to search tools to index the content and use it to provide their services. The court’s interpretation of the facts is interesting, as it means that coming to the networks world would not be neutral Instead, it would imply consent to search engines practices, which would already seem as customary though really recent.

In the same time, this approach may be seen contradictory to individual views on the exercise of copyright: when a lawsuit is brought by an author, it means in itself that this right owner is at least reluctant to the use of copyrighted works by a search tool!

A Protective Regime for Search Engines?
It is only a very limited number of persons who decided to dispute the possibility for search tools to use what they posted online. For the search engine companies, thus, the risk of being sued may seem extremely low. Nevertheless, it remains problematic that search engines may at any moment suffer from a literal application of the EU copyright regime, given the social significance of their activities.

The fact that European courts oscillate from one opinion to another reveals a lack of legal security. A change in the law or a ruling of ECJ (the Court of Justice of the European Union) in favor of implied licenses to unambiguously protect search engines in view of their social role, could be a way to find a balance between the necessary defense of intellectual property rights and the crucial access to information.

Cédric Manara is currently (2010) an affiliate researcher at the IPR University Center. His study summarized above was carried out with the help of a gift for research from Google Inc.