Google as publisher in Australia and the United Kingdom

Kate Ballis and Veronica Scott (Melbourne, Australia)*

Rapid advances in technology and a surge in the use of electronic communication have created new challenges for legal systems around the world.  One emerging area of law that courts are considering with more frequency is liability for online publication. In the last 12 months, courts in Australia and the United Kingdom have made findings that United States-based Google Inc (‘Google’) is the publisher of its Google Search results, Google Image results, Google AdWords, and its blog service, (‘Blogger service’). These findings suggest that internet service providers (ISPs) could be liable in Australia or the UK for defamatory content published on their services. The Supreme Court of Victoria has gone so far as to suggest that even ISPs that passively provide internet services could be a publisher of content for the purpose of defamation law.

Trkulja v Google Inc,1 Tamiz v Google,2 Google Inc v Australian Competition and Consumer Commission,3 and Rana v Google Australia Pty Ltd4 all provide a useful indication of the direction that common law courts are taking in relation to online publication. However, these decisions should not necessarily have the chilling effect that they initially appear to convey if ISPs effectively manage the risk of publishing defamatory content. Risk mitigation strategies are further discussed below.

Trkulja v Google (Google Image and Web Search results)

In November 2012, the Supreme Court of Victoria ordered that Google pay Milorad (Michael) Trkulja AU$200,000 after a jury trial. The claim related to material in Google Images and Google Web Search (‘Google Search’) results. Mr Trkulja claimed that the material conveyed various defamatory imputations about him by connecting him to criminal gangs in Melbourne.

The jury decided that Google had published the Google Images and Google Search results, that one of the three defamatory imputations alleged was conveyed by both the Google content, and that the defence of innocent dissemination was available to Google until the time it was put on notice of the defamatory material contained in the image search results by Mr Trkulja’s lawyers. As Mr Trkulja never put Google on notice of the material contained in the web search results, Google was entitled to rely on the defence of innocent dissemination for the entire period of publication.

When the jury returned its verdict in response to 24 questions, Google applied non-obstante to have the jury verdict dismissed and for judgment in its favour. As a result, Justice Beach needed to consider whether the jury was entitled to return the verdict they had given. He did. In delivering his reasons for dismissing Google’s application, he expressed the view that Google was liable for publication both before and after it was put on notice. Beach J found that it was open for the jury to conclude that Google was a publisher, even before it had any notice by Mr Trkulja because:

‘Google Inc intended to publish the material that its automated systems produced, because that was what they were designed to do upon a search request being typed into one of Google Inc’s search products.’5

Beach J relied on common law principles, and drew an analogy between Google Inc and a newsagent who sells a paper containing a defamatory article. While the newsagent may not have intended to publish the defamatory material, it has the relevant intention to publish the newspaper for the purposes of defamation law. Although a publisher, the newsagent could usually rely on the defence of innocent dissemination.

Beach J stated that the question of whether a particular internet service provider is a publisher in respect of defamatory content is ‘fact sensitive’6 and found that Mr Trkulja’s case could be distinguished from the three key decisions of Justice Eady of the High Court of England and Wales on ISP liability for defamation (Bunt v Tilley7 Metropolitan Schools Ltd v Designtechnica Corporation;8 and Tamiz v Google Inc9) in which he found that ISPs (including Google) were merely passive providers of the material and were not publishers. Beach J distinguished Bunt v Tilley and Tamiz based on their facts as they related to the hosting of internet services without intervention (Tamiz involved Google’s Blogger service) as opposed to Google Search and Image results. He noted that the facts in Metropolitan Schools are similar as they both relate to Google’s search results, but he considered that Eady J did not consider the fact that search engines do exactly as intended by those who own them and provide their services.

In any event, Beach J stated that none of these cases form the common law of Australia, because under Australian law, even if an entity’s role is passive, it can still be a publisher.10 In support of this proposition, Beach J footnotes the discussion in Urbanchich v Drummoyne Municipal Council11 which focuses on the lack of intent required and refers to Justice Hunt:

‘The law of defamation […] has never required ‘a conscious intent to induce the public or any individual to read the alleged libels’ […]The proposition that conduct of a passive nature cannot amount to publication […] was decisively rejected in Byrne v Dean […] Greene LJ said he was quite unable to accept any such proposition. Nor am I […]’12

Beach J found that on these facts the jury could find that Google had intended to publish the material that its automated systems produced, and further distinguishes Bunt v Tilley on the basis that an internet intermediary performs more than mere facilitation of the postings. However, he appears to consider that the common law in Australia does not require conscious intention to publish, but instead conduct of a passive nature can amount to publication. Although not making a decision on this point, Beach J certainly suggests that in Australia, an ISP could be a publisher even if it acts passively.

Tamiz v Google Inc (Blogger service)

In February 2013, the UK Court of Appeal in Tamiz overturned the decision of Eady J (referred to above) in relation to publication, and found that after notificationof the defamatory content, Google was publisher of the content. However, although Google was a publisher, the Court of Appeal found that there was no real or substantial tort, as it was highly improbable that a significant number of people would have read the comments on the blog between Google becoming liable (at some point after notification) and the blog being removed. Note that counsel for Tamiz raised the Trkulja decision as evidence that Australian courts have not accepted Eady J’s Bunt v Tilley analysis. However the Court of Appeal did not consider the decision. This suggests that courts in both UK and Australia have separately decided that Google is the publisher of content.

While Google was found to be a publisher after notification, the Court of Appeal did not consider that prior to notification Google was a primary or secondary publisher. The Court of Appeal did not consider that Google was a primary publisher (eg, author, newspaper proprietor) as it does not create the blogs or have any prior knowledge of, or effective control over their content. The Court of Appeal did not consider that Google was a secondary publisher (eg, distributor) because there was no evidence to show that Google knew or ought by the exercise of reasonable care, to have known that the publication was likely to be defamatory. Google would have no way of monitoring content on its Blogger service to determine whether content was defamatory beforeit was put on notice of such content.

As Google was not the publisher prior to notification, then beforenotification it could rely on the innocent dissemination defence in section 1 of the Defamation Act 1996 (UK). This defence protects persons who (a) were not the author, editor or publisher of the statement, (b) took reasonable care in relation to its publication and (c) did not know and had no reason to believe that they caused or contributed to the publication of a defamatory statement.

When considering whether or not Google was a publisher after notification, the Court of Appeal took a very different approach to Eady J. It relied on the principles in Byrne v Dean (also relied on by Beach J)to find that Google was in fact publisher of the defamatory content on the Blogger service. The Court of Appeal considered that in hosting the Blogger service, Google was analogous to the golf club that had a notice board on which a person posted an allegedly defamatory note and the golf club allowed it to remain there for days. At first instance, Eady J drew an analogy between Google hosting the Blogger service and ownership of a wall that was covered in graffiti. The Court of Appeal instead found that in providing the blog service, Google goes further than providing a notice board, as it provides tools to help the blogger design the layout of the blog and enables the blogger to display advertisements. In addition Google can readily remove or block access to any notice that does not comply with its terms of use that the user accepts. Earlier in the decision, the Court of Appeal stated that Eady J:

‘was wrong to regard Google Inc’s role in respect of Blogger blogs as a purely passive one […] by the Blogger service Google Inc provides a platform for blogs, together with design tools and, if required, a URL; it also provides a related service to enable the display of remunerative advertisements on a blog. It makes the Blogger service available on terms of its own choice and it can readily remove or block access to any blog that does not comply with those terms.’13

This sits at odds with the Trkulja decision in which Beach J distinguished the Blogger service from the web and image search results on the basis that the Blogger service was provided by Google without intervention. From the above description, Google appears to be as active in the provision of the Blogger service as it is in the provision (and creation of the program to create) the search and image results.

In Byrne v Dean, Greene LJ established that the test as to whether the golf club was liable for not removing the content was:

‘having regard to all the facts of the case is the proper inference that by not removing the defamatory matter the defendant really made himself responsible for its continued presences in the places where it had been put?’14

The Court of Appeal found that the facts fell within the scope of reasoning in Byrne v Dean and, if Google allowed defamatory content to remain on its Blogger service after notification:

‘it might be inferred to have associated itself with, or to have made itself responsible for, the continued presence of that material on the blog and thereby to have become a publisher of the material.’15

However, the Court of Appeal did not consider that such inference could be drawn until Google had been given areasonable timewithin which to act to remove the defamatory material. What is a reasonable time in general terms was not decided by the Court of Appeal. However, by taking five weeks to respond, was considered by the Court of Appeal to be too long, and sometime before then, Google was brought within the Byrne v Dean principles and became a publisher.

It is interesting to note that, although it gave such heavy weight to Byrne v Dean, the Court of Appeal did not consider Greene LJ’s finding in Byrne that a purely passive ISP could be liable.

In terms of whether the section 1 defence continued to apply immediately after notification,the Court found that Google was not a publisher (in the commercial sense of the definition in the Defamation Act, even if at some point after notification it became a publisher at common law), and that it had taken reasonable care, thereby satisfying requirements of section 1. However, the court found that at some point after notification Google knew or had reason to believe that its continued hosting of the material in question caused or contributed to the publication. At that point in time, Google could be liable for publication of defamatory content on its Blogger service. This appears from the court’s reasoning, to also be the point in time at which the Court of Appeal deemed Google to be a publisher at common law.

The effect of this decision is that if Google does not move quickly to take down defamatory content it has notice of, if enough people viewed the content before removal (or if the content was never removed), it could lose the protection of the section 1 defence and be liable for publication of the defamation.

Google Inc v ACCC (Google AdWords)

In February 2013, the High Court of Australia (Australia’s final court of appeal) found that Google was not responsible for misleading or deceptive conduct in relation to its Google AdWords links. In making their decision, the judges concluded that Google did not author the sponsored links, it merely ‘published or displayed’ them. While this may have helped Google avoid liability for misleading or deceptive conduct, this decision is relevant to defamation law, as is authority of the High Court that Google is publisher ofits AdWords content. The High Court found that:

‘The technology which lies behind the display of a sponsored link merely assembles information provided by others for the purpose of displaying advertisements directed to users of the Google search engine in their capacity as consumers of products and services. In this sense, Google is not relevantly different from other intermediaries, such as newspaper publishers (whether print or online) or broadcasters (whether radio, television or online), who publish, display or broadcast the advertisements of others. The fact that the provision of information via the internet will – because of the nature of the internet – necessarily involve a response to a request made by an internet user does not, without more, disturb the analogy between Google and other intermediaries.’16

This shows the direction that the High Court might take if it were to make a decision on whether an ISP was ‘publisher’ for the purpose of defamation law. The decision is also an indication of what the High Court may decide on the point of publication if Google were to appeal the Trkulja decision.

Rana v Google Australia Pty Ltd

In Rana v Google Australia Pty Ltd,17 the plaintiff claimed that Google Australia had control over its search engine and was therefore liable for defamatory content on websites in Google Search results. However, Rana failed against Google, with the court finding there was no reasonable prospect of establishing that Google Australia was the publisher of the websites hosted by Google Australia. Counsel for the defence submitted that:

‘Google Australia is not authorised to, and has no ability to, control or direct the conduct of Google Inc and is not responsible for the day-to-day operations of Google Inc; Google Inc owns and operates the domains and The search engines at the domains mentioned are exclusively provided by, operated by, and controlled by Google Inc; and Google Australia does not have any ability to control or direct action in respect of blocking URLs from’18

The Federal Court relied on the fact that in A v Google New Zealand Ltd,19 Google New Zealand obtained summary judgment on the basis it was the wrong defendant. The Court concluded in that case that the operation and control of the Google search engine resides with Google Inc and not with Google New Zealand.

The Federal Court also relied on the fact that in Tamiz v Google Inc and Google UK Ltd,20 Eady J held that Google UK Ltd simply carries on a sales support and marketing business within the UK and does not operate or control

Implications for ISPs

These authorities do not necessarily have the chilling effect they initially appear to convey.

From a UK perspective, Tamiz was an interlocutory application made by Mr Tamiz to serve Google Inc outside of the jurisdiction. Eady J held that the court should decline jurisdiction. Mr Tamiz appealed this decision to the Court of Appeal. The substantive claim has not yet gone to trial. As Beach J said in Trkulja, it is not possible to know whether the type of evidence that was presented to the jury at trial would have been available to Eady J in the interlocutory application or whether any new evidence was provided to the Court of Appeal. In Trkulja, Beach J’s decision was about whether it was open to the jury to make a finding on the evidence at trial. A different decision inTamiz could be made by the court at the trial phase.

In addition, in Tamiz, the Court of Appeal found that there was not a substantial tort. The question in the UK is therefore when content posted on a blog that was subsequently removed by the blog owner could amount to a substantial tort. This would depend on the extent of publication before the content was removed and the damage caused. Secondly, in the UK, it appears that if Google responds to takedown requests and removes defamatory content within a reasonable time, it may be protected. While (based on the Trkulja decision) in Australia, as soon as Google is put on notice it can no longer rely on the innocent dissemination defence; if Google maintains processes to effectively manage the risks of publishing defamatory content then liability can, in many cases be avoided.

While Beach J said that any decision on whether Google is a publisher of content is fact specific, he has also indicated that passive ISPs could be liable as publishers. The UK Court of Appeal’s decision appears, however, to distinguish the facts in Tamiz from those in Bunt v Tilley on the basis that Google’s role in the Blogger service is not a purely passive one. This suggests that an ISP in its most literal sense that passively provides an internet service, may not be liable based on the decision in Bunt v Tilley.

Beach J’s finding indicates that the decision in Australia could extend to a provider of internet services to a home or office, any online program developers, or App developers, that develop software with the intention of generating new material, and even to an ISP providing an internet service to a website. Again, it will be fact specific and Beach J stated that an ISP may not be found liable for every service that it provides.

In Victoria, if ISPs are liable as publishers of defamatory content, they may look to the publishers of the original defamatory content on the website – such as the website owner, online newspaper publisher or author of the article – for contribution under section 23B of the Victorian Wrongs Act 1958, as their websites are set to allow Google to crawl and include their content in its search results.

Of course, in the internet world, subject to being granted leave to serve outside the jurisdiction, the location of the ISP is irrelevant. If the content was downloaded in Australia or the UK and the person has a reputation in either country, then Google could potentially be liable. Whilst ISPs have a bit of room to move on the Tamiz facts and decision, things do not look as hopeful for ISPs that host content that is downloaded in Australia.

Risk mitigation

In order to prevent any chilling effect resulting from these three recent decisions, ISPs should develop an effective, user friendly, and clear and functional take down policy and procedure for defamatory content which users of the site can read and understand. This may include requiring the complainant to identify:

  • the source of the defamatory material, for example, search results, a webpage, a blog etc, and provide the URLs;
  • the specific content contained in the material that they consider defamatory;
  • the identity of the person(s) allegedly defamed;
  • the defamatory imputations that they believe arise from the content;
  • where the complainant is based;
  • when the information was first published;
  • any indicators of the extent of publication; and
  • any other information they think it is necessary for the ISP to know in order to make an informed decision.

This kind of information should assist the ISP to assess the risk of leaving the content online. It may require engaging with the complainant and seeking a legal view as to the risks of a claim for being successful in the jurisdiction in which the person resides. This may involve considering the seriousness of the nature of the material, whether the ISP could be a publisher, whether the content is simply reproduced or is a composite of sources and what defences it could rely on.

Where the ISP considers that the content is defamatory, then it should remove the content as soon as possible.


Given the direction some common law courts are taking, ISPs are facing the prospect of being found to have published defamatory material. Absent a blanket defence or safe harbour, the question of liability and defences will turn on the particular facts: the nature of the publication; notice of the defamatory content; and the ISP’s response. Risk mitigation therefore cannot be avoided. Effective take-down policies and procedures will be necessary.  


* Kate Ballis is a former-lawyer previously at Minter Ellison. Veronica Scott is Special Counsel at Minter Ellison. Veronica can be contacted in relation to this article at

1 [2012] VSC 533.
2 [2013] EWCA Civ 68.
3 [2013] HCA 1.
4 [2013] FCA 60.
5 Trkulja [18].
6 Trkulja v Google Inc and Google Australia Pty Ltd [2012] VSC 533 at [18].
7 [2007] 1 WLR 1243.
8 [2011] 1 WLR 1743.
9 [2012] EWHC 449.
10Trkulja [28].
11 (Unreported, Supreme Court of New South Wales, 22 December 1988).
12Trkulja, at footnote 41.
13Tamiz [23–24].
14 Ibid [29].
15 Ibid  [34].
16 [2013] HCA 1 [69].
17 [2013] FCA 60.
18 Ibid at [35].
19 [2012] NZHC 2352.
20 [2012] EWHC 449 (QB).