Trkulja III not as good as the originals: Google Inc v Trkulja [2016] VSCA 333

By 2008, Indiana Jones had fought, flown and fallen through several great adventures: all against great odds, in the face of powerful empires and despite significant hurdles, he ploughed on… ultimately to victory. Then came Indiana Jones and the Crystal Skull. One movie too many.

In 2012, Michael Trkulja famously scored defamation victories against Yahoo! and Google for publishing search results arising from searches of his name on their websites. Those results wrongly linked him to the Melbourne criminal underworld. Trkulja won $225,000 and $200,000 against these giants respectively.  For the internet giants, the Victorian Supreme Court really was the Temple of Doom.

Like Dr Jones, Trkulja felt he had another big win in him. He sued Google again. But like Dr Jones at the box office, the Court of Appeal concluded that this was just one adventure too many.

This time, Trkulja’s statement of claim asserted that he was defamed by:

  1. Compilations of thumbnail images returned by Google’s search engine in response to certain terms searched, usually related to the Melbourne criminal underworld (the Images Matter);
  2. Search results returned by the Google search engine to a range of searches that also involved the Melbourne criminal underworld (the Web Matter); and
  3. Autocomplete predictions that appeared at different stages of Trkulja’s name. Those predictions also involved the Melbourne criminal underworld.

Trkulja essentially claimed that these search results/ publications were defamatory of him because they carried imputations including that he was:

  • A hardened and serious criminal in Melbourne;
  • An associate of notorious criminals such as Andrew Veniamin, Carl Williams, Tony Mokbel; and
  • Such a significant figure in the Melbourne criminal underworld that events involving him are recorded on a website that chronicles crime in the Melbourne criminal underworld.
Victorian Supreme Court of Appeal

Victorian Supreme Court of Appeal

Earlier, Google had unsuccessfully applied to the Trial Division to have Trkulja’s case against it struck out. At that stage, Justice McDonald dismissed the application [2015] VSC 635.

Not to be denied, Google appealed. To win, it had to persuade the Court of Appeal that Trkulja had no real prospect of success.  It was a significant obstacle, but one that it ultimately met.

Google ran three arguments: firstly, that as a matter of law, it could not be held to have published the search results that Trkulja sued on, secondly that in any event, those search results were not defamatory of Trkulja and thirdly, that a search engine provider should be immune from a proceeding such as this. The last of these arguments was very ambitious and dismissed very quickly by the court. The first two arguments however, were dealt with in great detail and ultimately in Google’s favour.

The Images matter comprised image search results from the Google search engine. Those results were attached to the statement of claim and ran for 20 pages, mostly comprising thumbnail pictures. The search terms were a variety of terms that included Melbourne criminal underworld and variations to that theme. Of the many images displayed, an image of Trkulja was included, alongside many others, including specific Melbourne underworld figures.

However, the pictures also included a raft of apparently random people: the Chief Commissioner of

Marlon Brando

Marlon Brando

Police, a barrister, a solicitor, a murder victim, a journalist, Marlon Brando, earlier reports of other defamation proceedings instituted by Trkulja, the St Kilda pier and a Melbourne tram. Trkulja claimed that the search terms themselves, taken together with the thumbnail photos of himself alongside Melbourne underworld criminals, were defamatory of him. Justice McDonald had originally concluded that this was arguable.

As for the Web Matter, Trkulja’s statement of claim referred to a number of Google autocomplete predictions, if a person searched Trkulja’s name and a range of Google search results for terms like “Melbourne underworld criminals” and the like. Trkulja claimed that all of these Google results, including the autocomplete predictions, without more, were defamatory of him.

Trkulja gave notice of these complaints to Google on 3 December 2012 and asked Google to take them down. Notice is normally very important, because in previous cases, in order to establish publication, the plaintiff had argued that once Google received notice, if it did not remove the material within a reasonable time, it became a primary publisher of the material complained of.

Justice McDonald found that Trkulja had arguably been defamed and that it was a matter for trial.

The crux of Google’s appeal were its contentions that it could not be held to be a publisher of the matter complained of, and that Trkulja had no prospect of proving that the matter was defamatory of him.

Is Google a publisher of its search results?

The appellant

The appellant

Addressing publication at common law, the Court of Appeal laboured through its history. The critical difference was between a primary publisher and a secondary publisher. The Court noted that even if a person did not take part in the original chain of distribution, a person could still be held to be a primary publisher of material from the date when they were taken to have authorized or acquiesced in the earlier publication of that material by another [112]. This normally happened because the publisher was held to have acquiesced, ratified or taken responsibility for the publication. To establish this, a plaintiff normally had to show that the publisher had notice of the matter complained of, had the authority to remove that matter and did not do so.

The main example was the Byrne v Deane case when a letter was stuck to a noticeboard belonging to a golf club and officeholders at the golf club were sued as publishers because they knew about the letter being up, they had the power to remove it and they did not.

This was different to situations where a person intendedly, but unwittingly, distributed material within which defamatory matter appeared. In those instances, a person could be described as a secondary or subordinate publisher [113].  For a secondary publisher, it seemed that notice of the publication took less of a role and those publishers often relied on an innocent dissemination defence.

In this case, Trkulja ran the argument that Google was a primary publisher and that was all. That was critical. He did not argue that Google was a secondary publisher [225].

The Court of Appeal analysed many decisions relating to defamation claims brought against search engines, blogging sites, Facebook sites and Youtube sites, and noted that claims against site operators are to be treated differently to cases against search engines [285]. The question remained whether Google was a publisher of its search results. The Court concluded that the role of the search engine was not as a “passive instrument” [342] and at [348]-[349]:

[348] On first principles, we consider that a search engine, when it publishes search results in response to a user’s enquiry, should be accounted as a publisher of those results – and in this we include autocomplete predictions. It is a participant in a chain of distribution of material.”

[349] We also consider that a search engine should be accounted a secondary publisher….”

In all, the opinion of the Court was that Google was a secondary publisher of the search results and could avail itself of the innocent dissemination defence, at least before it was notified of the material and arguably after that too. It was not a primary publisher [357].

In this matter, the claim by Trkulja did not allege that Google was a secondary publisher, which the Court considered would have been arguable [370]. Trkulja only argued primary publication, and the Court concluded that this was not arguable. This meant that the way that Trkulja had put his case was not arguable and on that basis alone, the case ought to be dismissed.

Could the Images Matter or autocomplete predictions be defamatory?

Regardless of the Court’s conclusions on publication, these ultimately did not matter because the Court concluded that, in any event, the publications complained about were not capable of being defamatory [372].

The Court concluded that in determining meaning arising from the internet, the question must be determined by reference to the understanding of an ordinary reasonable user of a search engine such as Google [390].

On that basis, Trkulja would have no prospect of establishing that the Images matter conveyed any of the defamatory meanings asserted. Twenty pages of random images meant that the reader would not understand them in the way the plaintiff claimed. The plaintiff’s picture was simply one within a random compilation of images [400]. A reasonable user would be aware of the unpredictable results generated by an image search.

The Court concluded that the user of the internet, in context, would understand a disconnect between the images and the search terms and a repeat user would understand that the search results in their entirety did not reflect the meaning of the inputted words considered as a phrase [151].

The Court also concluded that autocomplete predictions could not be understood in a defamatory way [393], they are simply a collection of words that had been entered by previous searches.  Further, the other material sued upon were simply references to Trkulja’s previous defamation successes and the Court did not understand those to be defamatory. Overall, the matter sued on was incapable of being defamatory [412] – the case was dismissed.

So when the dust settles on all of this, unless Mr Trkulja seeks and obtains leave from the High Court, it looks like it might be his last case against an internet giant. Although one never knows. Indiana Jones 5 is apparently set for the screens in 2019…..

This entry was posted in Defamatory meaning, Publication and tagged , , , . Bookmark the permalink.

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>