In December 2021, the Australian Government released an Exposure Draft of its Social Media (Anti-Trolling) Bill (the Bill), promoting it as offering “some of the strongest powers to tackle online trolls in the world”.
The Bill purports to protect Australians who administer and maintain social media pages and provide for the ‘unmasking’ of anonymous online trolls. Unfortunately, if enacted in its current form, it is our view that the Bill will not assist in combatting online trolls and may in fact undermine aspects of defamation law which already go some way towards achieving this aim.
Even the title of the Bill is a misnomer, focusing solely on defamation law rather than the wider issue of ‘trolling’ – which often takes the form of abuse, bullying and harassment but which, whilst hurtful, may not be defamatory.
The Bill responds, at least in part, to the High Court’s September 2021 decision in the Voller defamation proceedings (see References at end).
That decision (which we discussed here) found that the operators of Facebook pages may be regarded as ‘publishers’ of any comments made by third parties on those pages. In response many media organisations sounded the tocsins – the decision, it was suggested, would unfairly ‘open the floodgates’ to unlimited online liability for the operators of Facebook and other social media pages, even in relation to posts written by others.
The existing law - following Voller but prior to the Bill - is that:
- The author of the comment (the ‘commenter’) is potentially liable as a publisher (that is, the effect of Voller is not that the operator of the social media page is the only publisher of a third party comment on that page; the commenter is also responsible as a publisher); and
- So too are those who operate the social media page on which the third party comment is posted (the ‘page operator’) as well as those who provide the social media service (the ‘provider’). This position operates as a significant incentive for page operators and providers to properly moderate the content on their pages and platforms respectively.
Traditionally, and prior to Voller, for a victim to proceed against an anonymous commenter (or ‘troll’), they could seek preliminary discovery from the platform (eg. Google or Facebook) to obtain the IP address of the commenter and if necessary from the ISP to which the IP address is assigned. All of this would result only in identifying the owner of the device registered to the IP address, and further steps would be required to prove the identity of the troll who used this device to post the defamatory material. Only then could the victim potentially initiate proceedings against the troll.
By contrast, following Voller and other recent decisions confirming the liability of big-tech platforms as publishers (see References at end), a victim of trolling may simply choose to sue the page operator or provider, which is likely to divulge the identity of the troll in the ordinary course of discovery, particularly as they have an incentive to do so to reduce their potential liability. Providers and page operators usually remove defamatory posts before reaching this point, making it a much simpler and more cost-effective route to identifying anonymous trolls than the traditional method described above. This would no longer be possible if the Bill was enacted.
Summary of the Bill
The Bill proposes to vary this position – by providing that an Australian person who administers or maintains a social media page is not a publisher of a third party comment posted on that page (see ss.3(a), 13(a),14(1)(c)). On the other hand, the provider is to be regarded as a publisher (see ss.3(b),13(b),14(1)(d)).
Further, the Bill would create a new conditional defence for the provider (under s.15), if it has in place and complies with a ‘complaints scheme’ which meets the ‘prescribed requirements’ (at s.16).
The complaints scheme
The Bill would allow a person (the complainant) to make a complaint to a provider about a comment posted by a third party (the commenter) on a page of the provider.
In summary, where a complaint is made, and provided the comment was made in Australia, the scheme would provide that:
- Within 72 hours, the provider is required:
- To inform the commenter about the complaint (s.16(1)(b));
- To notify the complainant that it has done so (s.16(1)(c)); and
- To provide the complainant with “country location data” (s.16(1)(d)), which is defined (at s.6) as meaning a statement as to whether the commenter appears to have been located in or outside of Australia when they posted the comment (according to the geolocation technology used by the provider)
- The provider may remove the comment from the page (but only, it seems, with the commenter’s consent) (s.16(1)(e));
- The provider must notify the complainant of the outcome of the handling of the complaint, within 72 hours of that outcome (s.16(f)); and
- If the complainant is dissatisfied with the outcome, it may then request that the provider disclose the commenter’s “relevant contact details”; the provider must ask if the commenter consents to the disclosure of those contact details; and, if the commenter does consent, the provider may disclose those contact details (s.16(g)).
The provider is not required to take any action at all if it reasonably believes that the complaint, or the request for the commenter’s contact details, does not genuinely relate to the potential institution of a defamation proceeding by the complainant against the commenter in relation to the comment (s.16(h)).
The Bill would also allow a prospective plaintiff to seek a court order against the provider for the commenter’s contact details (s.18).
An order may be obtained where:
- An Australian person (the prospective applicant) reasonably believes they may have a right to obtain relief, in a defamation proceeding against the commenter in relation to the comment; and
- The prospective applicant either is unable to ascertain the commenter’s contact details, or is unable to ascertain whether the comment was made in Australia, or reasonably believes the comment was made in Australia; an
- An Australian court would have jurisdiction if the prospective applicant were to institute a defamation proceeding.
The court would have discretion to refuse to make a disclosure order, including if it was satisfied that the disclosure could present a risk to the commenter’s safety (s.18(3)).
The power conferred on the court under s.18 is expressly “in addition to, and not instead of, any other powers of the court” (s.18(5)) but it is not obvious that it does add to the court’s existing powers. For example, the Federal Court has frequently made preliminary discovery orders requiring Google LLC to provide details that might identify the owners of Google accounts who, either anonymously or under pseudonyms, published disparaging Google reviews (see references at end).
New defence for the platforms
Section 15 of the Bill would provide platforms with a defence in relation to a third party comment where: the comment was made in Australia, the provider has a complaints scheme that meets the s.16 prescribed requirements, and either the complainant has not requested the commenter’s contact details and the court has not made a disclosure order, or, alternatively, if the complainant has requested the commenter’s contact details or the court has made a disclosure order, the provider has disclosed the relevant contact details.
Unfortunately, there are several fundamental problems with the Bill, to the point that, in our view, it should not proceed in its current form.
Immunity for page operators
Perhaps the most significant aspect of the Bill is that it confers a complete immunity on social media page operators (to that extent reversing Voller).
The operators of social media pages (such as Facebook pages) should, at least in some circumstances, be potentially liable for third party comments –for example, if they are actually aware a comment is defamatory and refuse to remove it.
For a victim of a defamatory Facebook post, for example, the cheapest and most efficient means for that person to preserve their reputation is usually a written request to the operator of the Facebook page to request the post’s removal. If it is not removed, the page operator will presumably be unable to rely upon the defence of innocent dissemination. The page operator is therefore incentivised to remove defamatory posts.
By contrast, the Bill would provide page operators with a blanket immunity, regardless of the seriousness of the defamatory post, even after they are put on notice of it, and even if they unreasonably refuse to remove it. With that sort of immunity, it would seem significantly less likely that page operators would monitor the pages they control responsibly and remove unlawful posts upon request. This does not appear to achieve the objective of protecting Australians from harm online.
Pursuing the commentators themselves
One of the stated objectives of the Bill “is to connect potential victims of defamatory statements to the originator of those comments for the purpose of enabling the victim to institute defamation proceedings in Australian courts”
It is open to question whether the Bill will be effective in that regard:
- The Bill does not oblige a provider to have a complaints scheme, though it aims to incentivise providers to have such a scheme (by making it a precondition for the defence under s.15).
- In the absence of a court order, all the provider is required to disclose (under the prescribed requirements for the complaints scheme) is a statement whether the commenter appears to have been located in or outside of Australia at the time of the post in question. The provider is not required, it seems, to take any steps to verify where the commenter was in fact located.
- Although promoted as being about ‘unmasking’ trolls, the Bill appears to contemplate that the provider would only reveal the commentator’s contact details if they consent (and why would they?). Even then, the provider is not required to provide the complainant with the relevant details (although it must do so to avail itself of the defence).
- There is no requirement that providers obtain, or take reasonable steps to obtain, accurate contact details in the first place, so any contact details which are disclosed may not be genuine (though again it appears the details must be genuine in order for the provider to have a defence under s.15).
- Any contact details that are disclosed might not be sufficiently precise, or reliable, to allow the complainant to initiate proceedings, rendering the complaints scheme useless.
- Even where the complainant obtains useful contact details (which can already be obtained by court order) and initiates proceedings against the commenter, the proceedings could well be expensive and lengthy. The Bill does nothing to assist with this.
- After all this, if the complainant successfully obtains orders against the commenter, these orders may be unable to be enforced, and there is no guarantee either that the existing post will be removed or that an injunction will be issued to restrain future posts.
- Perhaps most notably, there is no requirement for the platform to remove a defamatory third party post (except with the commenter’s consent). The requirement for consent imposes an obstacle to the removal of the post which does not currently exist – providers often unilaterally remove third party posts from their platforms in order to reduce their potential liability for them.
Accordingly, the victims of defamatory social media posts may be left without any remedy whatsoever. That is, they may have no remedy against the commenter (for the reasons just provided); they may have no remedy against the page owner (because they are given immunity); and they may have no remedy against the provider (because of the new defence).
Nor does the Bill do anything to streamline the complicated and expensive traditional method of identifying online trolls, including seeking preliminary discovery from big-tech platforms and, potentially, ISPs. Worse still, by potentially removing the liability of big-tech platforms, it removes the ability for victims of trolling to circumvent this complicated traditional method by suing the big-tech platforms themselves, who already have both an obligation and incentive to divulge the identity of the troll in the course of ordinary discovery in such a case.
For the reasons outlined about it is our view that the Bill should not proceed in its current form.
But if it is to proceed at all, the Bill should be considered as part of Stage 2 of the review of Australia’s Model Defamation Provisions (‘Review of MDPs’), which is being led by the States’ Attorneys General, and which has been underway for some time, with Stage 1 of the reforms implemented in several states last year.
A focus of the Stage 2 Review of MDPs is the decision in Voller. In our view, Voller and other recent decisions holding page administrators and big-tech platforms potentially liable as publishers of third party content has led to good policy outcomes. Fake news and cyber-bullying have pernicious consequences for society; it is appropriate to hold big-tech platforms in particular to account, as an offset to generating profit from user engagement online.
At the same time, there should be a policy objective to maintain the primacy of professional, responsible journalism over less scrupulous rivals. There are many options for reform under consideration in the Review of MDPs. One option may be to treat big-tech platforms as publishers of third party comments, but not page administrators. Another may be to enact the dissenting judgments of Steward and Edelman JJ in Voller to make page administrators liable only where something in their conduct can be shown to have ‘procured’ or ‘provoked’ the trolling. But to fundamentally alter defamation law at the very time that a detailed, stakeholder-driven and established process is ongoing is premature and, for the reasons set out above, ill-conceived.
There are undoubtedly areas of defamation law that require reform. A mandatory, streamlined take-down procedure for defamatory content on big-tech platforms, similar to the model operating effectively in the UK, is desirable. The impact of Voller on responsible media outlets is also worth considering, as is the enforceability of Australian court orders against foreign domiciled big-tech companies. Outside of defamation law, which is this Bill’s focus, more can be done to hold big-tech to account to protect against cyber-bullying and online trolls. This bill, despite its title, would unfortunately achieve nothing in that regard.
- Social Media (Anti-Trolling) Bill 2021: A Bill for an Act relating to defamatory material posted on social media services, and for related purposes
- Dylan Voller Defamation proceedings: Nationwide News Pty Ltd v Dylan Voller  HCA 27
- Fairfax Media Publications Pty Ltd v Dylan Voller  HCA 27
- Australian News Channel Pty Ltd v Dylan Voller  HCA 27
- Kabbabe v Google LLC  FCA 126
- Allison v Google LLC  FCA 186
- Kukulka v Google LLC  FCA 1229
- Seven Consulting Pty Ltd v Google LLC  FCA 203
- Sydney Criminal Lawyers v Google LLC  FCA 297
- Lin v Google LLC  FCA 1113