Worth a thousand words? Videos and the implied freedom of political communication in Farm Transparency v NSW
Henry Palmerlee
10.02.2022
Why are videos such a valuable medium for communicating a political message? This question has intrigued members of the High Court since the freedom of political communication was first recognised to be implied within the Constitution in Nationwide News Pty Ltd v Wills (1992) 177 CLR 1 and Australian Capital Television Pty Ltd v Commonwealth (No 2) (1992) 177 CLR 106 (‘ACTV’).
Justice Gageler’s dissenting judgement in the recent case of Farm Transparency International Ltd v New South Wales [2022] HCA 23 (‘Farm Transparency’) includes an insightful explanation of why video-based political communication has attracted such consistent judicial interest. His Honour argued that two unique characteristics of visual imagery – its persuasiveness and inherent veracity – make it a potent medium for political communication in the modern age. This analysis played an important role in Gageler J’s conclusion that ss 11 and 12 of the Surveillance Devices Act 2007 (NSW) (‘the Act’) impermissibly burdened the implied freedom.
In this piece, I consider Gageler J’s reasoning in the context of recent advances in video manipulation through artificial intelligence, specifically deepfakes. I argue that the advent of deepfakes heralds the arrival of an era in which we cannot always believe our own eyes. Contrary to the position advocated for by Gageler J, judges may no longer be able to rely on the supposed truth of video content as a relevant factor when considering the proportionality of legislation.
Visual imagery and political communication
The implied freedom of political communication recognises that the Constitution provides for a system of representative democracy in Australia, a system which requires free and open communication about political matters by the public. A common thread throughout multiple High Court decisions dealing with the implied freedom has been the Court’s interest in the visual communication of political issues.
This interest has been apparent since the birth of the implied freedom in ACTV. Both Mason CJ and McHugh J emphasised that the impugned provisions, which gave incumbent candidates for election preferential treatment for television advertising, restricted access to ‘the most effective’ political advertising medium (a point that has been affirmed as recently as in Brown v Tasmania (2017) 261 CLR 328, [258]). McHugh J extended this line of reasoning in Levy v Victoria (1997) 189 CLR 579, 624 (‘Levy’), reflecting that:
No one could fail to understand the impact of the war in Vietnam on the civilian population after seeing the picture of a terror-stricken, naked child running away from her burning village. Such an image probably had more to do with influencing United States public opinion against the war in Vietnam than any editorial of The New York Times or Washington Post … Without the opportunity to use the medium of television, the citizen cannot make use of its unique communicative powers.
Justice McHugh also accepted the appellant’s submission that visual depiction of cruelty has ‘a dramatic impact’ that differentiates it from other media of communication such as the written word. In his Honour’s estimation, the appellant capturing and publicising images of bloodied ducks to advocate against hunting was ‘more likely to attract public attention to their cause’ than other forms of communication (at 625). Notwithstanding that McHugh J ultimately upheld the impugned legislation, his Honour recognised this factor as militating strongly in favour of a finding of invalidity.
As political advertising shifts increasingly onto social media, the importance of visual communication in politics is only more apparent. Nearly 27,000 political ads were run on Facebook and Instagram in the two months preceding the 2022 federal election, underscoring the centrality of visual imagery in informing electors’ choices at the ballot box. In Farm Transparency, Gageler J once again considered the efficacy of visual communication in the context of political communication. His Honour’s reasoning demonstrates why visual communication has consistently attracted the interest of members of the High Court, and how consideration of the medium of communication is relevant to judicial scrutiny of legislation. Justice Gageler’s reasoning, however, introduces a new consideration to an analysis of the video-medium: its supposedly inherent veracity, or truthfulness. This consideration, I argue, ought to be questioned.
Video-based communication in Farm Transparency
In his Honour’s reasons, Gageler J pointed to two unique qualities that make video-based communication particularly appealing for political purposes: its persuasiveness, and its inherent veracity. Picking up on the analysis discussed above, Gageler J recognised that the ‘persuasive power of visual imagery’ means that this sort of communication has a particularly dramatic impact on its audience: [80]. Since it is such an effective means of persuading the voting public, visual communication is essential in the political arena. Such reasoning bears a resemblance to McHugh J's discussion of the video-medium in Levy outlined above.
However, beyond the quality of persuasiveness, Gageler J also laboured to make explicit that the inherent veracity of visual imagery is another factor which makes visual communication so important to those wishing to communicate a political message. His Honour variously described visual communication as ‘of its nature … factual’; a source of ‘true factual information’; ‘factual information … known to its possessor and potential communicator to be true’; and an ‘extant record of activities’ relevant to voters: [79]-[81]. Visual communication is not only valuable because of its effective emotional appeal, but also because it presents a true record of events. When we see a video of something – for example, animals being subjected to inhumane conditions – we know (or we can at least assume) that it has really taken place. In contrast to simply reading an exposé on a topic, video evidence provides incontrovertible proof that something has occurred, even if it is later edited and presented in a way that serves a certain agenda. Justice Gageler’s judgement in Farm Transparency therefore covers new ground in arguing that the inherent truthfulness of the video-medium is relevant to the judicial scrutiny of legislation.
Visual communication’s relevance to proportionality analysis
The ‘factual [and] peculiarly communicative’ nature of visual communication is relevant to the Court’s assessment of whether an impugned law impermissibly burdens the implied freedom of political communication: [80]. Following McCloy v New South Wales (2015) 257 CLR 178 (‘McCloy’), a structured proportionality approach is applied to determine whether a burden is impermissible. Farm Transparency is the first case since McCloy in which a member of the High Court has explicitly dealt with video-based communication, and Gageler J’s reasons give the clearest demonstration to date of the relevance of the medium of communication to scrutinising legislative provisions.
Although Gageler J is critical of the application of structured proportionality (see, eg, Brown v Tasmania (2017) 261 CLR 328 at [158]-[164]), his Honour’s analysis in Farm Transparency demonstrates how the inherent veracity of the video-medium would be relevant to whichever approach is preferred by a future High Court. On this matter, Gageler J stated stridently that:
The prohibitions are too blunt; their price is too high; the cost they impose on the communication and receipt of information about matters of political and governmental concern is more than could be warranted for every activity which might be shown by a visual record to have occurred on private property (emphasis added): [82].
In other words, the restrictions on political communication were not ‘adequate in their balance’: [82]. To restrict political actors from using such a potent medium of communication formed an impermissible burden on the implied freedom. A blanket ban on such an important source of ‘true factual information’, in Gageler J’s view, could not be upheld.
The impact of deepfakes
Justice Gageler’s analysis demonstrates how the supposed veracity of visual communication might be relevant to assessing the permissibility of a burden on the implied freedom. However, advances in artificial intelligence pose a fundamental threat to the logic underpinning this analysis. Deepfakes – manipulated videos that replace a person’s face with a computer-generated face – challenge the presumption that a video is ‘of its nature’ factual. These videos can be produced by anyone with a relatively powerful computer, and can be used to do anything from doctoring a video of Morgan Freeman to spoofing Nicholas Cage’s face onto Amy Adams’ body. Deepfakes can be extremely convincing, and are becoming more and more difficult to detect even with the assistance of computers. Under the guise of the video's apparent objectivity, deepfakes can be used as a tool to distort the past. As such, I would respectfully express some caution about both the current relevance and future applicability of Gageler J’s statement that videos invariably contain ‘true factual information’ that requires strict protection.
It might be argued that videos have never actually presented a true record of events. We could point to techniques like selective editing, in which video clips are spliced, shortened or otherwise edited to present a particular narrative. A politician’s speech might be manipulated, or an interaction shown out of context, to lend support to a message or point that the editor wishes to communicate. On this view, Gageler J’s reasoning regarding the inherent veracity of videos might have in fact been outdated since someone first spliced two sections of film together.
This argument is certainly persuasive. However, there is a fundamental difference between selective editing and staging compared to deepfaking: only deepfakes are capable of producing a convincing video of something that never actually occurred. While the previous example of a politician’s manipulated speech can be edited to appear in a certain way, it can only ever rely on existing material. In contrast, the possible material for use in a deepfake is endless, as a deepfake can be created completely from scratch, limited only by the creator’s imagination.
Indeed, deepfakes have already made their entry onto the scene of political communication. Climate advocacy group Extinction Rebellion deployed deepfake technology to great effect when they released a deepfake video depicting Belgian Prime Minister Sophie Wilmès discussing climate change’s connection to COVID-19. Other prominent political figures such as Vladimir Putin, Barack Obama and Donald Trump have all been the subject of widely circulated deepfake videos. In the past, we could safely assume that a video of a politician discussing policy was, at the very least, something that had occurred – whether or not it was a completely objective representation of their platform. Deepfakes undermine this assumption. Whether used for comedic purposes, to make a policy point, or to pursue a more nefarious aim, deepfakes are becoming a fixture of the political arena.
I argue, therefore, that deepfakes must be seen as the guillotine blow to any assumption that videos present a true record of events. It may be true that videos have never functioned simply as passive vessels of history, immune to manipulation or partisan agendas. This is the case with most source material. However, until recently, a video still required a subject at which to aim the lens and press record. This requirement no longer exists. Deepfakes are completely fabricated, resembling what French semantic philosopher Jean Baudrillard referred to as ‘simulacra’: signs which bear ‘no relation to any reality whatsoever’. Simulacra are reflections and distortions of other unreal signs, like an endless hall of mirrors. In the context of deepfakes and political advertisement, Baudrillard’s argument that ‘only TV functions as a medium’ for the communication of these unreal simulacra appears prescient. Video is a potent medium for the communication of simulacra, and political advertisers have already realised the potential offered by this new technology.
While it may have previously been wise to express some scepticism about the inherent truthfulness of the video-medium, it now seems imperative. Deepfakes are unique in being completely divorced from reality, able to present synthetic depictions of events that never occurred. They go far beyond manipulation; deepfakes are not based on creative editing or convincing staging, but on no truth whatsoever. It therefore appears that Gageler J’s emphasis on the inherent veracity of the video-medium may be outdated. The future of video manipulation has arrived, and proportionality analysis must adapt to recognise this.
Where to for video communication?
If courts are now faced with a serious challenge to the assumption that videos are inherently true records of events, how might they approach laws restricting video communication in the future? I argue that deepfake technology will likely result in courts placing a reduced emphasis on visual communication as distinct from other media. Through deepfakes, video has definitively lost its status as an indisputable record of truth.
However, I will first address one way in which deepfakes are unlikely to influence the implied freedom of political communication: by being ruled out as a mode of political communication. Andrew Ray has convincingly argued that deepfakes constitute political communication for constitutional purposes, for example as satire or parody. Past cases have demonstrated that political communication may appear in many forms, and that analysis will centre on the substance of the communication rather than the form it takes. So long as a piece of communication ‘might be relevant’ to, or ‘could affect choices in federal elections’, it is at least arguable that the communication is relevantly political (Lange v Australian Broadcasting Corporation (1997) 189 CLR 520, 561). As such, laws regulating deepfakes will likely still be subject to the implied freedom of political communication.
Instead, it appears that courts must recognise that the inherent veracity of videos is a thing of the past. This does not mean that there is now nothing to differentiate videos from other media when assessing legislation that governs visual communication. Justice McHugh’s comments in Levy – that the emotional persuasion power of visual communication is completely different to the written word – remain salient. However, the explicit importance placed by Gageler J on videos as a reliable source of ‘true factual information’ may well fall to the wayside. In the same way a satirical article or photoshopped image can carry a political message without containing ‘true’ content, political communication through video is no longer an inherently truthful source of information.
Additionally, the veracity of the video-medium may become a factual battleground between litigants. When assessing the adequacy of an impugned provision that regulates video communication, a court might be required to decide a question of fact as to whether videos are inherently truthful. As Gageler J’s reasoning demonstrates, regulation of such a medium of communication may well require a higher level of justification to be adequate in its balance than a law controlling a medium which does not possess the same quality of veracity. In arguing for or against the truthfulness of the video-medium, parties to future cases may have to lead expert testimony on whether deepfakes can be detected, what sort of videos they can produce, and how convincing they are to a public audience. Far from reducing the importance of the video-medium to implied freedom jurisprudence, deepfakes are likely to give rise to new and intriguing issues in cases to come.
Although videos will undoubtedly continue to have enduring relevance to Australian politics, it seems that their inherent veracity can no longer be assumed in the context of proportionality analysis. This fundamental re-evaluation of the extent to which videos are reliable evidence will have wide-ranging consequences beyond constitutional law. Will video evidence continue to be accepted in criminal trials? What civil remedy would be available if a deepfake of an ASX100 CEO caused the company’s stock price to plummet? As future cases regarding the freedom are litigated – especially with pressure mounting on the Federal Parliament to pass truth in political advertising legislation – the place of videos in our political process, real or fake, will surely be a central issue.
—
Henry Palmerlee is a Tuckwell Scholar at the Australian National University, completing a Bachelor of Laws (Hons) specialising in Public Law. He is also a member of the editorial board of the Federal Law Review.
Suggested citation: Henry Palmerlee, ‘Worth a thousand words? Videos and the implied freedom of political communication in Farm Transparency v NSW’ on AUSPUBLAW (10 February 2023)<https://auspublaw.org/blog/2023/2/worth-a-thousand-words-videos-and-the- implied-freedom-of-political-communication-in-farm-transparency-v-nsw/>