We live in an era where use of technology including Twitter and Facebook has become the norm for most organisations. But what should businesses do when negative comments are posted on their pages, and who can ultimately be held liable for such actions?

The recent publicity around the practice of ‘trolling’, where individuals deliberately post inflammatory, provocative or derogatory comments, raises a number of interesting questions not only for the major social networking sites, but also for any organisation that hosts its own online forum that allows users or members of the public to leave comments.

While history tells us there are no easy answers, this article deals with the potential minefield that can ensue when a third party makes defamatory comments online.


Organisations must be aware that all “publishers” of defamatory material, including owners of a website, can be liable for defamation.

Although there are a range of defences available to any claim for defamation, a defence often relied upon for online publishers is that of “innocent dissemination”. This defence applies to protect publishers where they have no ability to read or vet everything they publish prior to sale or publication. Generally, to successfully rely on this defence, an online publisher must establish:

  1. that it was not the primary publisher or distributor of the matter; 
  2. that it neither knew, nor ought reasonably to have known, that the matter was defamatory; and 
  3. the lack of knowledge was not due to any negligence on their part.

The Australian position: Trkulja decisions

The Australian courts are yet to deliver a judgment which categorically determines whether a social media site is liable for defamatory content created by their users. However, when it comes to internet search engines like Google and Yahoo!, the Victorian Supreme Court has confirmed they may be liable for defamatory search results and image ‘linking’, even where the results are the product of automated algorithms.

In November 2012, Justice Beach awarded $200,000 damages to Mr Milorad (Michael) Trkulja, after he sued Google Australia and Google Inc for defamatory search results and image linking. The search results falsely suggested that Mr Trkulja was, amongst other things, a prominent figure in the Melbourne criminal underworld.

On discovering the search results, Mr Trkulja wrote to Google requesting that it remove the content and amend the search results. Despite being put on notice of the publication and having the power to remove it, Google failed to act, leading to the inference (based on the facts of that case) that it consented to the publication.

The Supreme Court confirmed that where such an inference is able to be drawn, an internet search engine may (in the same way as a web forum host) be liable as a publisher of defamatory content regardless of the fact that entity did not, intentionally or otherwise, create the defamatory material. Notably, Mr Trkulja also sued Yahoo! and received a similar judgment in that case.

More recently in 2014, an eBay seller sought to sue a West Australian man over a negative review he posted on the auction website. While this case was ultimately dismissed on a procedural point by the New South Wales District Court, it serves as a timely reminder that even something as common as customer feedback can, if not properly managed, lead to defamation proceedings.

Practical tips for Australian organisations

Publishers of web forums may wish to consider the following tips to help reduce the risk of being found liable for another author’s comments.

  • If it is not practicable to review and vet posts before they go live, implement a process for investigating and, if appropriate, removing any posts that other users validly object to. 
  • Act promptly to respond to complaints about posts. 
  • Prevent users from posting anonymously, or if that is not possible or desirable, ensure that the forum administrator has access to verifiable user information. 
  • Prevent users from posting comments without first signing up to appropriate terms of use. Preferably this will include a term allowing disclosure of the user’s true identity (if aliases are permitted) to any user who claims to have been defamed by that user’s post. 
  • Be careful about “featuring” or otherwise republishing the comments of any third party user without exercising a degree of editorial control and fact checking.