By Mitch Clarke
Consider how many electronic Internet links you click each day on your mobile or laptop. You presumably clicked on a hyperlink to arrive at this very article. The Internet and linked content are reciprocally essential; the benefits of one cannot be realised without the other. Their invention and use has advanced how we communicate, share content, and find information. Prohibiting the use of linking content would be antithetical to the Internet. Unfortunately, Trkulja v Google [2017] HCATrans 129 could have this effect.
Granted special leave in June, Trkulja presents two questions to the High Court. The first is procedural, predominantly as a result of the plaintiff initially being self-represented. The second is about whether a person who collated third party posts and linked-content on the Internet can be held liable if the content is defamatory. Generally speaking, the contentious issues in Trkulja and similar cases involve the liability of an Internet company for hyperlinking, collecting, collating, or reproducing content posted by third parties on the Internet.
There was a possibility that the second question would not be addressed and instead referred back to the Victorian Court of Appeal. I use past tense (‘was a possibility’) because of the October decision on a similar question from the Full Court of the South Australian Supreme Court in Google v Duffy. The Duffy judgment begs — and hopefully necessitates — a more critical adjudication from the High Court on the question of indexed and re-communicated Internet content. Without comment, there could be an unfortunate impairment — however unintended — on the operation of the Internet in Australia.
Precedent Spells Trouble for ‘Publication’
A defamation claim requires three elements: a ‘publication’, ‘identification’ of a third party, and ‘defamatory’ content about that third party. Without the presence of each element, there is no claim to be made. Much of the Australian case law has focused on the specificity of information reproduced on a webpage via hyperlinks, and the inclusive degree of such needed to constitute the re-publication of defamatory content underlying the hyperlink. However, it is the operation of hyperlinks and the associated reproduced content constituting a ‘publication’ which is deserving of more scrutiny.
‘Publication’ is not defined in the uniform legislation which is how we wind up with common law principled cases like Trkulja. Instead of developing principles conducive to present-day technology and innovation, courts have been ‘trying to fit a square archaic peg into the hexagonal hole of modernity’. Due to the State-level coordination required to unify defamation laws, the last reforms took 30 years to be enacted from the initial tabling of a Law Reform Commission report. The technical legal principle applicable to Internet companies which has been developed by the common law is called ‘publication by omission’. This means that the Internet company doesn’t become liable until after they have been notified about the defamatory material on their site and given reasonable time to remove it. Both domestically and internationally, judges have scrutinised the degree of control and intention an Internet company has on ‘publications’ that are automated or posted by users of the platform; but the outcome from the judgments have been vastly different. Australian principles have been derived from case law where a person could be held liable for defamatory material posted on their property by a third party after they became aware of the material and did not remove it. Principles regarding hyperlinks then developed from 19th century case law which dictates that pointing to a defamatory sign can be a publication.
Even when lower courts have tried to develop principles analogous to the historical case law so as to suit to the reality of the Internet’s place in society, appeal courts have overturned the findings, instead strictly conforming to analogies which are not suitable for the use of technology. The leading judgment from the Chief Justice of South Australia in Google Inc v Duffy evidences a similar progressive unwillingness (at [173], emphasis added):
the hyperlink, if used, will direct the searcher to [the intended] material. A Google search paragraph is the electronic analogue of the person who places a post-it note on a book which reads ‘go to page 56 to read interesting [potentially defamatory] gossip about X’. This approach also sits more comfortably with the numerous ways in which a hyperlink might be constructed and thereby refer a user to the underlying page.
An initial view of this analogy might find it to be correct, but it falls apart fairly quickly with a more critical reflection. For example, Google’s content indexing capability would equate to a person who has the ability to point all information online all at once. They can also account almost instantaneously for changes to that information. It is wrong to analogise this capability with a person who has the ability to post only a minute fraction of the same references (post-it notes) and must undertake analogue reviews to update the referral text on the post-it note. Additionally, the information provided by a company like Google is unique depending on the specific search terms of the user, but also upon the psychological profile of that user built over a period of time to more accurately predict the correct results of the user’s search. As much is noted by Google in its submissions in Trkulja.
Presently, digitally posting (as well as sharing or sending) content can attract liability for the person who posted it, in addition to the Internet company whose platform the link is posted on. This is true for the Internet company or re-posting party even if the content behind the hyperlink becomes defamatory only after the posting of the link (that is, it is retrospectively altered by the original author). As the law progresses in this direction, the fairly obvious result will be Internet companies restricting platform use — in part or in whole — within Australia, out of fear of defamation lawsuits. Indeed, this possible ‘chilling’ effect on digital communications has motivated courts in other countries to indemnify any liability for Internet companies arising from a user’s content. Trkulja will be the first opportunity for the High Court to articulate the various principles developed by subordinate courts; and an international comparison points fairly clearly to where these scattered judgments have faulted.
A Technologically Conscious Judiciary
Trkulja is an example of the comparatively anomalous Australian 21st century questions of law that arise because of the absence of a Bill of Rights. In the US, the mere concept of hyperlinks being defamatory is untenable in light of individual rights of freedom of speech. In Canada, instead of prevailing individual rights, the Supreme Court held that Internet communications (and hyperlinks specifically) were an essential characteristic of the collective expressionistic rights of citizens and could therefore not be considered in and of themselves, a ‘publication’; and therefore, not defamatory. The Internet was an inseparable component of communication and speech and therefore any restriction on its operation would impinge upon those rights (at [34]):
The Internet’s capacity to disseminate information has been described by this Court as ‘one of the great innovations of the information age’ whose ‘use should be facilitated rather than discouraged’. Hyperlinks, in particular, are an indispensable part of its operation.
Australian courts have rejected a similar conclusion because of the absence of express conflicting rights. But I disagree entirely, and it seems more apt to say that judicial attitudes — or perhaps understanding — towards technology is the true barrier to more thorough critiques in Australian courts.
Interestingly, in other countries courts have been explicitly conscious (at [42]) of not spelling out unnecessary legal principles in their judgments which may adversely affect the operation of future innovative technologies. Conversely, as evidenced by the Victorian Court of Appeal in Trkulja, Australian courts have found it necessary to author 200 page judgments on principles restricting the use of technology when there was no need for such an opinion to be given. The case concluded in Google’s favour on procedural questions, but the vast majority of the extensive judgment focused on addressing legal principles affecting technological use — even when the parties had not made evidence submissions on that question.
When discussing control, intent, and knowledge in Duffy, the South Australian Supreme Court distinguished that ‘approval of the presence of the material, and not [approval] of its message’ is the requirement (at [132]). This is an extremely low threshold. They concluded that ‘[p]ractically, the drawing of attention to a defamatory statement cannot occur without the person knowing, or being in a position to know, of the presence of the defamatory material in what is being pointed out’ (at [133]). This archaic statement serves only as a refusal to adopt legal principles conducive to the present day. While the statement is rooted in earlier case law, it is fairly clear that for present day purposes, Google cannot possibly or practically be aware of every single piece of content; or often even the context of the imputations behind it after notification. This type of reasoning provides yet another compelling reason for the High Court in Trkulja to reverse these findings, even absent the broader free speech protections afforded in other countries.
In its submissions to the High Court, Google contends that the ‘common law should recognise that the operator of a search engine has immunity in respect of actions for defamation founded on automatically returned materials’. It is unfortunate that they provided nothing further to evidence why that proposition should be so, but I do agree.
The Court in Duffy rejected this same submission as antithetical to freedom of expression, noting that this would ‘result in the conferral of immunity upon search engine operators from suit for defamation’ (at [622]). Conversely, the Canadian Supreme Court considered granting this immunity to be necessitated by the risk to societal communications. Hinton J in Duffy considered only impinging upon individual speech rights (of which there are almost none in Australia, in this context and more generally): ‘I very much doubt … a chilling effect on the freedom of expression. If I am wrong, it is in my view for the legislature to intervene.’ (at [624]).
A Possible Escape: Objects of the Legislation
The common law may be the source of these less than ideal legal developments, but it would be advantageous if the High Court relied on the legislation to create a more favourable environment for Internet companies. The uniform defamation legislation is meant to exist concurrently with the general law tort of defamation. However, if general law principles need to be modified or abolished by explicit or necessary implication of a provision, the legislation will prevail.
Two of the Act’s objects are quite relevant for this purpose. The objects of the Act include not placing ‘unreasonable limits on freedom of expression and, in particular, on the publication and discussion of matters of public interest and importance’ (emphasis added), and the promotion of ‘speedy and non-litigious methods of resolving disputes about the publication of defamatory matter’.
Cases in other counties have often involved an assessment of ‘public interest’ when subordinating the operation of the defamation tort on the Internet. However, in Australia the assessment of ‘interest’ or ‘public interest’ has only been judicially adjudicated in relation to the verbiage of defences. There is little agreement on the proper characterisation or scope of what the ‘interest’ evaluation might be. When discussing the concept of ‘interest’ in relation to qualified privilege (a defence to defamation), the High Court in Howe v Lees [1910] HCA 67 has said:
The interest must be definite; it may be direct or indirect, but it must not be vague or insubstantial — so long as the interest is of so tangible a nature that it is expedient to protect it for the common convenience and welfare of society.
Additionally, ‘interest’ in the statutory context is even broader than that at general law. With that in mind, consider that the court is also required to ensure that statutory interpretation exercises to discover the definition of ‘publication’ are amenable to these legislative objectives; and, ‘interest’ as part of the defence should be given the same meaning as the objects of the Act. In this context, it is difficult to see how the Duffy judgment in relation to the assessment of individual rights as opposed to society at large, would be correct. Properly assessed, the interest is an interest of facilitating freedom of expression broadly and not on an individual basis. Immunity granted to Internet companies like Google, in my mind, seem like the only feasible outcome if the operation of law is going to be consistent with the objects of ‘public interest’ and ‘freedom of expression’. To operate otherwise would be to ignore the Act’s objects and to restrict the expressionistic importance that the Internet has come to so readily facilitate in society.
The characterisation of Internet companies as information access facilitators may provide further support. In the context of the Act’s objects, these companies are merely assisting the promotion of speedy and non-litigious resolutions. As a result of services like Google, individuals are more easily capable of finding defamatory information and therefore the companies are a facilitating mechanism for speedy dispute resolution. One could then conclude that an Internet company’s liability for these ‘alleged’ ‘publications’ would again be to subordinate the Act’s objects. This is yet another consideration yet to be drawn by Australian courts but prominent in other countries.
There was clear legislative intent to facilitate the ‘rapid growth of technology, particularly in the world of media, cross-border communications and the advent of the Internet’. Unfortunately, presently-binding cases have the opposite effect. Legal issues aside, the law’s current state hinders international digital commerce, and discourages technological innovation in Australia. It would be laudable if the High Court re-tooled the legal principles applicable to Internet companies with explicit consciousness of the role of the Internet within society, thus restoring a state of law more congruent with international counterparts. Otherwise, chalk this one up as another example that freedom of expression is a regrettably narrow concept in Australian law.
AGLC3 Citation: Mitch Clarke, ‘Courting Communication Anachronisms: Trkulja v Google [2017] HCATrans 129’ on Opinions on High (30 November 2017) <https://blogs.unimelb.edu.au/opinionsonhigh/2017/11/30/clarke-trkulja
Mitch Clarke is a 2017 graduand of the JD program at Melbourne Law School and works at legal tech company Xakia Technologies.