Are search results defamatory? Google’s not feeling lucky

Sticks and stones may break your bones, but search results can hurt you? At least, that seems to be way the Australian courts have been leaning. In the latest in a series of cases against Google, Google was unsuccessful in its bid to set aside a writ and statement of claim alleging defamation by their search results.

The person bringing the action was Milorad Trkulja, a veteran at bringing – and winning – defamation actions against Google.

In this instance (Trkulja v Google Inc [2015] VSC 635), Trkulja alleged that Google published defamatory material about him through their search results. A photo of him would come up when using Google Images to search terms such as ‘Melbourne underworld criminals’ and ‘Melbourne underworld killings’, alongside images of notorious Melbourne criminal figures. These pictures also came up when searching for such terms in the normal search functionality as a preview for the Google Images results.  Further, when searching for his name, Google’s autocomplete functionality would suggest the terms ‘criminal’, ‘Melbourne crime’ and ‘underworld’.

Trkulja alleged these results carried a number of defamatory imputations, such as:

  • Trkulja is a hardened and serious criminal in Melbourne in the same league as the other convicted criminals displayed;
  • Trkulja is an associate of the other convicted criminals displayed; and
  • Trkulja is such a significant figure in the Melbourne criminal underworld that events involving him are recorded on a website that chronicles crime in Melbourne criminal underworld.

(Not exactly the impression you’d usually want to give off…)

Searching for success

When asking to set aside the action, Google had to prove that the action had no real prospect of success. Google based its application on three main grounds, all of which Justice McDonald roundly rejected.

  1. Firstly, Google claimed that it could not be liable as a publisher as it had a passive, facilitator role as a search engine. Google argued there was no human input in the application of the search engine, and therefore it had no intention to publish the actual search results. However, Justice McDonald held that it was arguable that Google would be found to be a publisher, following the 2012 judgment of Beach J in a previous proceeding between Trjulka and Google. Importantly, Justice McDonald found that Google was capable of being a publisher both before and after notification of the alleged defamatory material.  Justice McDonald pointed to the fact that highly skilled programmers were employed to develop Google’s algorithms, and thus “the linking of Trkulja’s name with images of members of Melbourne’s criminal underworld is no coincidence but a direct consequence of the operation of the search engine in the way in which it was intended to operate.”  Justice McDonald also found that the Google submissions conflated the two separate questions of whether Google had published defamatory material and whether the defendant could avail itself of the defence of innocent dissemination.
  2. Secondly, Justice McDonald dismissed Google’s argument that Trkulja had no real prospect of establishing the results were defamatory. To be defamatory, it must be something which tends to lower a person’s reputation in the eyes of a reasonable person in the community. Justice McDonald found it was certainly arguable that a reasonable person would look at the compilation of images from the search results and assume that Trkulja was a convicted criminal like the other people depicted.
  3. Thirdly, Google submitted it would be in the public interest to grant an immunity from suit. Justice McDonald refused to grant an “immunity out of thin air”, noting there was no basis for such an immunity. In any case, he stated it would be up to the legislature, not the courts, to come up with any new defences or immunities.

Justice McDonald therefore dismissed Google’s application to set aside service of the writ and statement of claim on the basis that it had no real prospect of success.

The proceedings are continuing, and it remains to be seen what findings are made on these issues on a final basis.

A tangled web

The question of whether and in what circumstances a search engine will be liable for the content of search results is certainly a murky area of law. In Bleyer v Google Inc [2014] NSWSC 897, Justice McCallum of the NSW Supreme Court found the performance of the function of an algorithm was not capable of establishing Google’s liability at common law as a publisher of its search results, though the position after Google received notice of the defamatory material was “more complex”. Therefore in Bleyer, Justice McCallum expressly disagreed with the conclusion reached in the earlier 2012 Trkulja proceedings. However, as noted above, in the latest Trkulja decision, Justice McDonald expressly disagreed with Justice McCallum, finding that as Google’s search results were the outcome of the operation of the algorithms created by its employees, Google would not be prevented from being considered a publisher of those results.

The recent Trkulja decision also considered the South Australian judgment recently handed down in Duffy v Google Inc [2015] SASC 170, where it was found that Google was liable for defamatory content of its search results and autocomplete functionality, which linked Dr Duffy’s name to the term ‘psychic stalker’. Unlike Trkulja and Bleyer, which were both interlocutory proceedings, Justice Blue’s judgment in Duffy was a final decision. In that case, Justice Blue found that Google was a secondary publisher after it had received notice of the defamatory material and failed to remove it within a reasonable time. Google was recently ordered to pay Dr Duffy a whopping $100,000 in damages as a result of Justice Blue’s findings.

There are many questions that remain unclear following these first instance decisions. Will the line of authority in Bleyer be followed, or will Trkulja and Duffy be followed? Will a search engine be liable only once it has received notice of defamatory material and failed to take that material down?

At the very least, these decisions demonstrate that Australian courts may be willing in certain circumstances to find search results defamatory. This clearly has serious ramifications for search engines operating within our borders. Should Google monitor and edit its search results to avoid defamation suits such as this one? Alongside the recent ‘right to be forgotten’ ruling in Europe, this raises some tricky issues about the balance between privacy and freedom of the internet. Although Google’s argument about a public interest immunity was found not to have any legal basis, there are no doubt questions about whether liability really is best placed on intermediaries such as a search engine.

Google Maps can’t give direction for where this all is heading.