Sign in / Join

The liability of search engines: The Google case

As is well known, Google has two features that are used daily by millions of people, and namely the autocomplete and related searches functions.

When we search on the internet for the name of a person, the search engine suggests some related terms, which are sometimes offensive or defamatory.

Google - in interlocutory proceedings brought before the Court of Milan - argued that  such search is conducted by its software on the basis of a mathematical algorithm, whose role is merely passive.

Google invoked, therefore, articles 15 and 16 of Italian Legislative Decree No . 70 of 2003 concerning the activities of caching (which consists in the transmission on a communications network of information provided by a recipient of the service, so that the service provider is not liable for the automatic, intermediate and temporary storage of such information), and hosting (which consists in the storage of information provided by a recipient of the service).

Preliminarily, the Court stated that the possibility of exemption from liability provided for under Legislative Decree 70 /2003 should be regarded as being an exception to the rule and, therefore, should be strictly applied (on account of it being in contrast with the general Italian tort liability system).

This having been said, the Court referred to several previous decisions (i.e. a Milan Court interim order dated September 9, 2011 and a Rome Court interim order dated December 13, 2011 ) in which it was held that the technical developments in the field of Internet services have, in some cases, led to the traditional figure of ISPs (in which they are regarded as mere providers of computer technical support, which allows users to access the internet) being superseded, with the result that some service providers are - as regards the management of content by users - no longer completely passive and neutral figures (i.e. active hosting) and are actively involved in the organization and selection of material submitted by users, which  puts them in a position of no longer being in a the position of mere suppliers of storage space or of content or software which is viewed by third parties.

In the present case, the Court of Milan found that Google did not just passively host other people's content, but carried out additional activities which were not automatically and necessarily only related to the transmission or collection of content (such as the indexing, organization and selection of content), with the result that it had lost its passive and neutral status and had assumed an active role on the account of the fact that, even though its role could not be assimilated to that of a content provider (since it did not produce and transmit editorial content), the role played by it in interfering and organizing content was clearly not compatible with the neutral status provided for under articles 15 and 16 of the Legislative Decree 70 of 2003 .

The Court also held: The software used by Google is only theoretically "neutral " (on account of it being based on a system of automated mathematical algorithms) since the neutrality thereof comes amiss when - as a result of the application of such automatic algorithm based on criteria chosen by its creator - an improper combination between search terms arises therefrom. Nor can the mere fact that the system’s operating mode (software) creates the contested combination in an automatic fashion lead to it being considered “neutral" (cfr. judgment handed down by the Court of Milan on March 24, 2011).

Among other things, the services and features described above were, in the Court of Milan’s words, far from being essential for the provision of services and the storage of content and are to be considered additional features that enhance the Google search engine, making it more interesting and obviously appealing than less well equipped search engines.

In conclusion, it was found that the activity of so-called active ISPs, such as Google, should be examined in the light of the ordinary criteria established for tortious liability.

The judgment under review was also noteworthy for the comments made on the findings of the trial judge.

The Supreme Court held that, as regards the first of the plaintiffs (who was a physical person), the combination of the words cult and mind control with an individual’s name was not defamatory since this association can, at the very most, induce users to connect the person in question with the terms “cult” and “mind control”, without providing any further information in relation thereto. As a result thereof, we may (even where we consider such terms in a decidedly negative sense) attribute to such person either an active role or a passive role (as a victim of a cult or of an episode of mind control).

In the absence of a clear meaning being attributed thereto, the Supreme Court held that the words sect and mind control were not immediately related to the plaintiff and were, therefore, not offensive or defamatory. The interlocutory action had, therefore, to be dismissed with particular regard to the combinations of the words cult and mind control.

The Supreme Court held, on the other hand, that a completely different conclusion had to be reached when considering the association - made by Google’s autocomplete function - of the word cult with the name of the other plaintiff (which was a Foundation).

In this sense, the combination of the word Foundation with the word cult allowed users to immediately associate the concept of cult with such plaintiff and attribute to the latter the status of a cult.

The Court held that the aforementioned term had a clearly negative and derogatory  connotation, which conveyed information that was held to harm the reputation of the plaintiff (which was an organization with charitable and scientific purposes: The plaintiff Foundation claimed, in fact, to be a charity active in the field of scientific exploration, the promotion of social welfare and defense of children’s rights, as well as being involved in various international projects and in raising funds through donations).

According to the Court of Milan, it was quite likely that the harming of the Foundation’s reputation would have immediate repercussions on the economic support received by the latter from third parties.

There were, therefore, reasons of urgency which justified the immediate removal of the above-described combination operated by the Google’s autocomplete function.

The Court, in rejecting some of Google’s defensive arguments, held that  this Court believes that - contrary to what has been claimed by Google - there is no evidence to suggest that, in general, the users of the search engine have a level of knowledge of the Internet and its services that allows them to understand the fact that the suggestions proposed by "Autocomplete" are just the byproduct of a system that collects the most popular search terms, thus preventing them from being influenced in the way they judge the plaintiff (as a result of the combinations of words suggested by "Autocomplete")…It is, on the contrary, quite evident the rather negative impression of the plaintiff that is made on users when they read the combination of words in question. This impression is all the more striking on account of the fact that "Autocomplete" automatically completes the typing of the words in question, anticipating and suggesting terms that users might otherwise never have even taken into consideration.

This post is also available in: Italian

Leave a reply

*

Stanco dei soliti modelli di contratto trovati Online? Scopri Yourcontract!
Inizia ora
close-image

Utilizziamo cookie, anche analitici e di profilazione di terze parti, per migliorare l’esperienza di navigazione. Proseguendo con la navigazione acconsenti al loro uso in conformità alla nostra Cookie Policy. Regola le preferenze in Impostazioni Cookie