The High Court has delivered final word on search engine liability for third-party postings, in its Wednesday decision in Google LLC v Defteros  HCA 27 – at least for now.
The High Court upheld, by a 5-2 majority, that search engines are not ‘publishers’ of defamatory content by providing links to defamatory web pages.
Mr. Defteros is defamed
In summary, Mr Defteros, who is a Victorian criminal lawyer, alleged that he was defamed by search results returned by Google LLC in response to a search for his name. He was particularly concerned about:
- a 2004 article published by age titled “Underworld Loses Valuable Friend at Court”. Mr. Defteros initiated a procedure on this subject in 2016 (Procedure 2016); and
- other material, including the Wikipedia article “Melbourne Gangland Killings” which included a footnote to the article Underworld (Procedure 2017).
Mr Defteros argued that Google’s publication of the Underworld article defamed him on the grounds that Google was involved in the search engine function and therefore should be considered a publisher for the purposes of Australian defamation laws. Google became aware of the defamatory article in February 2016, when Defteros Lawyers made a removal request, but did not remove it until December of that year. During this period, the article was viewed 150 times.
On May 6, 2020, the trial judge ruled in favor of Mr. Defteros and awarded $40,000 in damages in the 2016 proceeding. The 2017 proceeding was later dismissed. Contrary to Google’s arguments, Justice Richards concluded that the search engine was not a “passive tool” and noted that identified objectionable content can be removed by human intervention from displayed search results.
Judge Richards also found the Underworld article to be defamatory, with the charge being that Mr Defteros had “crossed the line from professional lawyer to, to confidant and friend of criminal elements”. Google appealed the decision.
On appeal, the trial judge’s findings were upheld, following which Google appealed to the High Court.
Four Reasons Why Google Said It Wasn’t a Publisher
Google raised four issues on appeal:
- first, that he was not the “publisher” of material within the meaning of defamation law. Instead, it has simply provided a hyperlink to the material – the operator of the website to which the hyperlink relates is the relevant publisher of the material;
- second, that it should enjoy the qualified defense of privilege at common law on the basis that it provided an “indispensable means by which users can find information of interest to them on the Internet”. A substantial portion of the users to whom the Underworld article (which Google said concerned matters of considerable public interest) was published had a legitimate interest in it;
- thirdly, that it should benefit from the defense of qualified privilege under section 30 of the Defamation Act 2005 (Vic), because the material allegedly posted concerned a matter of real or apparent interest, and Google had acted reasonably in providing the hyperlinks; and
- fourth, Google was merely an innocent broadcaster of the material in question and should therefore benefit from the defense of innocent broadcasting.
The High Court decision: The search engine as librarian
By a majority of 5 to 2, the High Court upheld Google’s appeal based on its first point of appeal.
Chief Justice Kiefel and Justice Gleeson, with whom Justice Gaegler agreed, found that Google had in fact not published the material. In doing so, they went through the principles surrounding publication in defamation cases and likened Google’s involvement – that is, centralizing links in response to a search query – to someone asking for directions to a point of sale for a periodical containing defamatory material. In these circumstances, the Court noted that it would be “difficult to suggest that the person communicated the defamatory material contained in the periodical”. This was very different from the analogy made by the trial judge, that Google was a “librarian” handing someone a book marked with a particular page. Here, the Court said:
“The person is not directed to a particular outcome, as the Court of Appeal suggested. According to the findings of the trial judge, the results are ranked by the use of a relevance algorithm using “signals” or clues as to what the searcher is looking for depending on the words used in their inquiry. . The search result simply refers, in the sense of drawing attention, to a web page.
A search result was described by the Court as a reference to “something, somewhere else”. Facilitating a person’s access to the content of another web page was not considered to be participating in the two-way process of communicating its content to that person.
The Court also notes Mr. Defteros’ argument that, by not removing the material when informed of its existence, Google substantially acquiesced in the publication of the material and therefore “published”. This position was rejected – notice of the existence of the material was irrelevant to the question of whether the material had been published.
Judge Edelman and Judge Steward, in a separate judgment, also held that Google was not a publisher. In doing so, they agreed with Chief Justice Kiefel and Justice Gleeson that Google’s system was merely an indication that something existed in a particular location and that Google was therefore merely a facilitator rather than a publisher. Google also only provided information in the hyperlink that was provided by the website to which the hyperlink directed. He provided no other comments or color, defamatory or otherwise. Such involvement was not deemed sufficient to demonstrate that Google was in fact a publisher.
Justice Keane and Justice Gordon disagreed with the majority opinion, in separate opinions. Justice Keane held that Google was a publisher, in that its search engine helped people understand defamatory material. Publication to users – that is, their viewing of the material – would not have taken place without Google’s intervention, except in the rare case where a person can directly access the page via its URL, and therefore the search engine could not simply be described as a “telephone directory” or any other passive instrument. Judge Keane, however, did not express any detailed opinion as to Google’s liability for the publication or the availability of any defence, beyond merely confirming that he agreed with the Court of Appeals decision. on these issues.
Judge Gordon also held that Google was a publisher and pointed to the sophistication of Google’s crawler systems and the indexing of pages, including (and in particular) news pages. In his view, this, together with the commercial advantage that Google obtained through the exploitation of the search engine, provided a clear refutation of any claim that Google was a “passive” participant in the dissemination of the material to its audience. Judge Gordon also dismissed Google’s other grounds, finding that Google had failed to provide sufficient evidence that the material had been published to an audience interested in the material and therefore could not establish the defense of qualified privilege, and that couldn’t be an innocent broadcaster. in circumstances where he was made aware of the defamatory nature of the material and took no action.
Key takeaways for defamatory material, search engines – and anyone who hyperlinks
While the outcome of this case is welcome for Google, the lack of a unanimous High Court front means that a future case involving search engines could take heart if there is a differently constituted court.
The High Court’s decision is understandable in circumstances where the search engine can demonstrate that it was truly neutral (or passive, to use the term applied liberally by Google during the case). As noted in the judgment by Chief Justice Kiefel and Justice Gleeson, Google has been characterized as having acted as a tour guide for people interested in the larger topic that gave rise to the article, but does not did not compel readers to click on the hyperlink. It simply indicated that there was a website with material on that particular topic.
The situation could, however, be quite different if it were shown that a search engine has a particular bias or weighting towards articles with clearly defamatory tendencies. It depends on the systems and algorithms used by the search engine. With this in mind, the respective dissents of Justice Keane and Justice Gordon, and various comments made by the majority, suggest that the door may remain ajar for an aggrieved party in these circumstances to push for a different decision.
As noted above, the recent introduction in Queensland, NSW, Victoria and SA of a single publication rule and serious harm threshold may also serve to mitigate the potentially significant consequences of alleged defamation for some publishers. The next round of reforms will examine the liability of internet intermediaries for third-party content and potential protections for people who make reports to police, regulators and other relevant bodies in relation to alleged illegal activity. We therefore suspect that this could still be an additional layer of regulatory response to the issue of intermediary liability that has arisen in this case.
As always, it is wise and economically sound to obtain legal advice from an experienced defamation lawyer before making a decision to publish material containing or linked to material that could damage a person’s reputation.