Is Google News legal? (Updated)

In May, 2011 a Belgium appellant court released an important decision finding Google liable for copyright infringement by offering its Google News service to subscribers in Belgium at the and websites. The decision in Google Inc v Copiepresse et al, Brussels Court of Appeal (9th Chamber) May 5, 2011 suggests that services such as Google News cannot be legally offered in many countries around the world even from servers located in North America.

On February 14, 2012, the UK Copyright Tribunal in the Meltwater Holding BV v The Newspaper Licensing Agency Limited case suggested Google News and the Google Alerts service would also be illegal in the UK without a license from publishers.

The Google Inc v Copiepresse lawsuit involves Belgian press publishers, their Belgian authors and journalists (represented by their Belgian rights management companies) who had published press articles on sites operated in Belgium, the electronic address of which ended in “.be”. These articles were targeted at readers in Belgium. The dispute also involves Google which offers Google News using servers located in the US. The plaintiffs claim Google is liable for copyright infringement by copying articles from Belgium websites and disseminating the title and extracts of the articles from its Google.beand websites, without having obtained the prior authorisation of the publishers and authors.

The decision is lengthy and canvasses many issues including

  • the choice of law to apply to evaluate the claims of infringement,
  • whether Google’s caching of news articles is infringement,
  • whether the transmission of article titles and short extracts violates the reproduction and communication to the public rights,
  • whether Google News violates the moral rights of authors,
  • whether statutory fair dealing defenses apply,
  • whether Google can rely on implied licenses from publishers and authors to publish Google News, and
  • whether intermediary safe harbors are available under Belgium law for providers of these services.

While news about the decision was widely disseminated after its release, the far reaching holdings in the case were not as readily discernable. Perhaps this is because the decision is in French and no English translation was released when the decision was published.

The following is a summary of the main reasons in the case.

Google claimed that US copyright law was applicable to determine its liability as that was the place it cached the articles and from which it transmitted article titles and extracts to Belgium readers. The authors and publishers claimed the copyright laws of Belgium should apply to all claimed acts of infringement because Belgium had the closest connection to the acts of infringement.

The Belgium court held that all of the infringing acts were subject to Belgium copyright law. The court held that the infringement consisted of a set of complex acts which were located in different countries (the United States for the caching and transmission of query results and Belgium where the article titles and extracts were accessed). It ruled that this required the applicable law to be that of the law of the country most closely connected with these acts, which it said was Belgium.

In reaching this conclusion the Court relied on several authorities including the prior ruling of the Paris Court of Appeal in the La société Des Auteurs des Arts Visuels et de L’image Fixe Visual Auteurs (SAIF) v Google France  S.A.R.L. and Google Inc case. That case held that French copyright laws applied to Google’s image search service because that was the country in which, according to Article 5.2 of the Berne Convention, copyright protection was sought. In essence, the Paris Court held that Google’s acts of indexing and caching visual works of French authors and transmitting them to France from California were subject to the copyright laws of France. The Belgium court reached the same conclusion by holding that Google’s acts of indexing and caching news articles of Belgium authors and publishers and transmitting them to Belgium from California are subject to the copyright laws of Belgium.

The Belgium Court, like the Paris Court, might have concluded that the acts of indexing and caching which involve acts of reproduction that take place wholly in California should be analyzed under US copyright law and that the acts of communication to the public to persons located in Belgium should be determined in accordance with Belgium law. But, it did not do so.

A central issue in the case was whether Google infringed copyright by reproducing the title and extracts of news articles. In response to a user search, Google presented to users an article title taken from a news agency. It also copied the first three lines of the article and other titles on the same subject (by copying only the titles), an indication of the publisher and of the number of articles dealing with the same subject.

The Court ruled that this copying amounted to a reproduction of a significant portion of the publishers’ articles.[1] It held that Google could not reproduce the titles and excerpts from the articles without having first obtained their authorisation. It also ruled that the transmissions of this content to the public in Belgium infringed the communication to the public right.

The Court rejected various defences raised by Google including that the acts were permitted under fair dealing exceptions for the purposes of criticism, discussion, review, instruction, or scientific work. The Court also ruled that these exceptions would have to be read to be compliant with the Three Step Test which was carried into EU law by virtue of Article 5.5 of the EU Copyright Directive.[2] These exceptions could not be met as the reproductions in Google’s cache interfered with the normal exploitation of the works because internet users were able to gain access to articles that had either been removed from the publishers’ sites or were only available from the sites for a fee. Further, the Google News service deprived rights holders of reasonable compensation for new publications of their works.

The Court also rejected Google’s claim that it had an implied license to offer its service because the publishers did not elect to opt out of Google News by using appropriate website meta-tags. The Court also ruled that none of the safe harbours applicable to information society providers in the EU E-Commerce Directive applied to Google in the circumstances.

Google argued that the users of Google News were responsible for the acts of reproduction and communication, not Google. It contended that it only provided users facilities which an enabled these acts and so was exempt from infringement under the EU Copyright Directive. The Court rejected this argument holding that Google’s caching on its servers constituted a physical act of reproduction for which it alone was responsible.

As in Canada, under Belgium law authors have the moral rights of paternity and integrity. The Court ruled that both these rights were violated. The Google Newsservice infringed the right of paternity because the authors’ names were not mentioned in the search results. The right of integrity was infringed because only extracts of articles were reproduced and the works had been modified. The Court ruled that even though an Internet user knew, or should know, that what he or she was viewing was only an extract from an article and that it was possible to read it in its entirety by clicking on the hyperlink, that did not imply that the author had given his or her consent that only excerpts from the work could be published.

Google has the right to appeal the decision to the Cour de Cassation, Belgium’s highest court. Such an appeal appears likely given the importance of the issues. Accordingly to IPKat, a knowledgeable observer, it is also possible that Google may find a way to engineer a reference to the European Court of Justice either to buy time or generate a favorable ruling.

The Newspaper Licensing Agency Ltd & Ors. v Meltwater Holding BV case canvassed whether an online article clipping service can legally be operated in the UK without a license. Both the UK High Court and the Court of Appeal held it could not. After the Court of Appeal held that copying headlines and short extracts in providing the clipping service infringed copyright, the UK Copyright Tribunal was tasked with fixing a tariff for the use of the portions of the publishers’ articles that were copied.

In the course of the proceeding, Meltwater contended its business was similar to Google’s Google News and Google Alert services which paid nothing to publishers. It argued that it should likewise have no obligation to make any payments. The tribunal rejected this argument and in the course of doing so strongly suggested that Google’s services would also need to be licensed in order to be legally carried on in the UK. According to the tribunal:

There was some argument about the precise scope and effect of s129 on facts of this case. The point was as follows. Section 129 contains the word “shall”. The passage after sub-paragraph (b) states that the Tribunal shall exercise its powers so as to secure no unreasonable discrimination in certain circumstances. Such language is generally regarded as mandatory. The question is whether this mandatory language applies to the situation as between Meltwater and Google. Google must have licences from newspaper publishers to permit them to offer the Google News and Google Alerts service. However these licences are not part of the scheme and are not ones to which this reference relates. Google are not licensed under the WDL/WEUL. Indeed Google are not licensed by the NLA at all. The licences (assumed to exist) are directly from publishers….

The Google News and Google Alerts services are very similar to the Meltwater News  service save of course that it is not paid for by the end user. The articles are identified automatically without human intervention. Unlike Meltwater, Google do not provide  a personal service assisting with the search algorithms. Like Meltwater, the link provided by Google is a link to the newspaper’s website and behaves in the same way (subject to caching (below)). Both Meltwater and Google provide other related services as well and there was some unsatisfactory evidence about a further Google service called Google Reader. This arose as part of an argument about just how similar Meltwater’s overall offering may or may not be to Google. It seemed to us to be of peripheral relevance and we will ignore Google Reader.

We have addressed the basic services provided above. All three categories of service provider systematically monitor the media for clients and provide links to the newspapers’ articles. That is the core business and it is the same for all three of Google, Meltwater and the PCAs. Meltwater sought to distinguish themselves from both Google and the PCAs in that Meltwater only send links to articles whereas both Google and the PCAs can send whole articles to their users. This alleged distinction is more apparent than real. The argument about whole articles from Google related to internet caching by Google. It is a similar arrangement to a PCA’s internal database of scanned articles. The cached articles are only seen after a user has been sent a link and clicks on it to view the article. It is true that a PCA can send a whole paper copy of an article but by and large today what is sent by a PCA is a link – to one of the databases or to the newspaper’s website itself.

We find that all three services are based essentially on the same thing – identifying relevant articles and sending links to those articles to clients.

In the course of the decision, the tribunal noted that the Newspaper Licensing Agency had a mandate to collect royalties from Google for its Google News and Google Alert Services by commercial users, although it had not yet done so.

Before the High Court the NLA had stated that it had no mandate to license users of Google News. However it emerged before us that the NLA has been mandated to license systematic use of Google News / Google Alerts by commercial end users. Although it has not yet done so, the NLA intends to seek to require licences from commercial end users of Google Alerts who forward on the emails received within their businesses. Meltwater pointed out that this act of forwarding is additional to the core act of receiving the Google Alert service (or performing an ad hoc search). On any view this illustrates the significance of the Google News / Google Alerts service.

The Associated Press also just sued Meltwater in the US claiming copyright infringement and “hot news” misappropriation. This follows an earlier suit brought by several Canadian newspaper publishers in Ontario against Meltwater back in June 2011.

[1] Other courts have examined whether titles to articles can be copied. See, The Newspaper Licensing Agency Ltd & Ors v Meltwater Holding BV & Ors [2010] EWHC 3099 (Ch) (26 November 2010)  affirmed [2011] EWCA Civ 890 (27 July 2011) where a UK court held that the reproduction of titles alone was a reproduction of a substantial part of an article; and Fairfax Media Publications Pty Limited v. Reed International Books Australia Pty Limited [2010] FCA 984 in which an Australian court declined to protect titles to articles because they are insubstantial and too short to qualify for copyright protection as literary works and because the public interest requires using them for bibliographic reference purposes.

[2] Article 5 states: The exceptions and limitations provided for in paragraphs 1, 2, 3 and 4 shall only be applied in certain special cases which do not conflict with a normal exploitation of the work or other subject-matter and do not unreasonably prejudice the legitimate interests of the rightholder.“

Print Friendly, PDF & Email

One thought on “Is Google News legal? (Updated)”

  1. M says:

    The case was kind of funny because if you followed the aftermath.

    Google argued that rights-holders have a standard way of conveying their intentions using the robots standard. For whatever reason, Belgium’s court didn’t buy that.

    So they ordered Google to remove references to the Belgium news services from their site. Basically treat the site as a “deny=all”. Which Google did.

    But the newspapers quickly called Google “abusive” for complying with the wording of the court order, because without Google sending traffic to their site they were starting to lose money. Ouch!

    Basically for Google it is damned if you do, damned if you don’t.

    So eventually the newspapers agreed to use the robots standard to convey what they want their “rights” to be, which is what pretty much everyone else does who doesn’t want to be on Google News.

    Which is what they could have done from the start instead of suing Google. Technical solution instead of a legal solution? Unbelievable!

    Basically, a bunch of copyright lawyers made a bunch of money and nothing else really was accomplished in the end. But that’s not really a bad thing if you are a copyright lawyer now is it? 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Criminal copyright convictions of The Pirate Bay operators “necessity in democratic society” says human rights courtCriminal copyright convictions of The Pirate Bay operators “necessity in democratic society” says human rights court

The European Court of Human Rights (ECHR) has ruled that the convictions of Fredrik NEIJ and Peter SUNDE KOLMISOPPI, operators of The Pirate Bay bittorrent site did not violate Article 10 of the European Convention ...

%d bloggers like this: