When auto-complete goes wrong: High Court gives green light to defamation action for Google search results

Search engines have, without question, made access to information much easier and faster. Indeed, without search engines, it would be very difficult to navigate the trillions of internet pages and find meaningful results expeditiously. The algorithms behind search engines like Google are also now so developed that they are able to predict what you are looking for through their auto-complete function.  What happens though, if those functions return results that are claimed to be untrue and potentially defamatory?

In the recent decision of Trkulja v Google Inc [2018] HCA 25, the High Court of Australia allowed an action to continue in which Google Inc. (a US company) is alleged to be liable in defamation for publishing search results that include images of Mr Trkulja mixed with images of convicted Melbourne criminals, as well as text referring to him and predictions generated by Google's autocomplete functionality. The decision holds the door open to defamation proceedings. While it clearly affects search engines like Google, it may also have implications for businesses using search engine optimisation to increase their online profile (eg media providers where key words could be lumped together in an unintentional manner).

Relevant defamation law

The law of defamation provides a remedy where a person's reputation is damaged by the publication of unjustifiable derogatory information.

In order to sue, it is necessary to prove that the defamatory material that is complained of was in fact published by the party being sued. In order to prove that fact, it must be shown that the alleged publisher:

• was in some degree an accessory to the communication of the material in issue; and
• intentionally participated in the communication of the allegedly defamatory material.

As to whether something is actually defamatory depends on what ordinary reasonable people would understand by the matter complained of, and whether it would cause them to think less of the person in question.

The events leading up to the Google defamation case

Milorad "Michael" Trkulja is an Australian resident who was shot in the back during a shooting in  a Melbourne restaurant in 2004. This incident led to several articles in which there were references to certain Melbourne crime figures and investigations.

Mr Trkulja alleges that when Google image searches were performed during 2012 - 2014 of "Melbourne criminal underworld photos" and "Melbourne underworld criminals", images of him were mixed with images of convicted Melbourne criminals, and the pages contained various phrases such as "melbourne criminals", "melbourne criminal underworld figure", "melbourne criminal underworld photos", "Melbourne underworld crime" etc. It was also alleged that searches of Michael Trkulja's name associated him through the autocomplete function with terms like "is a former hit man", "criminal" and "underworld".

When Mr Trkulja issued proceedings in the Victorian Supreme Court, Google responded by applying to have the proceeding summarily dismissed on three bases:

• first, that it did not publish the images;
• second, that the matters in issue were not defamatory of Mr Trkulja; and
• third, that Google was entitled to immunity from suit as a matter of public interest.

The Supreme Court dismissed Google's application, finding that:

• by intentionally participating in the communication of the allegedly defamatory search results, there was a basis for alleging that Google was the publisher of the images;
• the fact that the images were often returned alongside well-known underworld figures meant that it was arguable that the material was defamatory as it suggested that Mr Trkulja was a convicted criminal; and
• the immunity proposed by Google was not in the public interest.

Google appealed the Court's decision on all three bases. The Court of Appeal did not decide the first ground (but noted its view that the innocent dissemination defence would likely be available), and rejected the third ground. However, it upheld Google's contention that Mr Trkulja would have no prospect of success in claiming that the matters in issue were capable of being defamatory. 

Mr Trkulja disagreed and appealed to the High Court.

Appealing to the High Court

The primary issue before the High Court was whether or not the defamation proceedings should have been summarily dismissed by the Court of Appeal.

In a joint judgment, the High Court upheld Mr Trkulja's appeal, finding that:

• it was strongly arguable that Google's intentional participation in the communication of the allegedly defamatory results to Google search engine users supports a finding that Google 'published' the allegedly defamatory results; and
• some of the search results complained of had the capacity to convey to any ordinary reasonable person viewing the results that Mr Trkulja was somehow associated with the Melbourne criminal underworld and those search results could accordingly be defamatory.

The High Court said that there could be no certainty as to the nature and extent of Google's involvement in the compilation and publication of its search engine results until after discovery. This adversely impacts a court's ability to make a summary determination on publication and innocent dissemination contentions.

As to whether the published material was defamatory, the High Court noted that the test for whether a published matter is capable of being defamatory is what ordinary reasonable people would understand by the matter complained of, and whether that would cause them to think less of a person. The test is not, as stated by the Court of Appeal, whether "any of the defamatory imputations which are pleaded [are] arguably conveyed". It is not what the Court thinks the allegedly defamatory words or images say or depict, but rather what they could convey to the ordinary reasonable person who is a member of the jury. This also must be considered in light of each search and response being, in effect, a different publication.

Further, while the High Court was prepared to accept the assumption that the ordinary reasonable person who has used the Google search engine contemplates that their search results bear some connection to the search terms, it cautioned that in the absence of tested, accepted evidence to the contrary, there could be a significant variance of experience and understanding among the range of persons taken to be representative of the hypothetical ordinary reasonable person

In short, the High Court held that the Court of Appeal had applied the wrong test in considering whether any of the pleaded defamatory imputations were arguably conveyed, and whether the case should be summarily dismissed. In those circumstances, the Court allowed Mr Trkulja's appeal and his case can continue before the Victorian Supreme Court.

Key takeaways from the Google defamation case

This case illustrates that search results (eg images that place a person in an incorrect context or allow false or improper connections to be drawn) have the capacity to convey defamatory imputations. Whether this case proceeds to create significant implications for search engines like Google, will depend upon whether the defamatory imputations are upheld and more importantly, whether Google is subsequently found to be the primary publisher or has a defence as an innocent disseminator under section 32 of the Defamation Act (Vic). No doubt, there will be many interested persons watching the progress of this case because of the impact it may have on how businesses use search engines.

It must be kept in mind, however, that the primary relevance of the High Court's decision focuses upon Google's application for summary dismissal of the claim and the standards and tests that must be applied. It demonstrates that the courts apply great caution when asked on an interlocutory basis to find that a defamation pleading is incapable of bearing a defamatory imputation.