The next speaker in this panel at the AANZCA 2025 conference is my QUT colleague Dan Angus, focussing especially on political advertising during the 2025 Australian federal election. This work is also supported by the ARC Centre of Excellence for Automated Decision-Making and Society. Computational advertising is ephemeral and targeted, individually personalised to the social media user; it is difficult to study these processes at scale. While platforms purport to provide some ad transparency libraries, these are limited, and can be enhanced through other approaches.
Some such approaches include data donations via browser plugins that capture the ads encountered by users as they use such platforms; another approach is mobile screen recording via an app installed on participants phones; the MOAT tool used for this study advances this further by ignoring any organic content and capturing only the ads that are shown to the user, thereby improving participant privacy and data gathering ethics.
The present project compensated its 100+ participants for installing and using this tool during the 2025 federal election. The data gathered by the tool are then reported back to a central server, where they can be aggregated and analysed. Such data can then also be correlated and compared with tools that connect to platform ad libraries, such as the PoliDashboard tool which does so for Facebook advertising.
Analysis of ad library information for the election campaign showed that (supposedly) unaffiliated advocacy groups spent the greatest amount of money on Facebook; the Coalition and Labor Party were the next largest groups. On Google, the Trumpet of Patriots party spent the most, followed by Labor and the Coalition.
Of the ads captured by the MOAT tool, only a very small number were political: 2% on Instagram, 5% on Facebook, and virtually none on TikTok. Most of these occurred during the final days of the campaign, and the focus was especially on Queensland – Brisbane, Griffith, Moreton, and other electorates were especially prominent.
Many such ads were attack ads or self-aggrandising; many made inflated and misleading claims, and often focussed on the key political leaders. Energy was a key campaign issue, with third-party ads from groups such as ‘Mums for Nuclear’ especially spruiking nuclear and gas energy. Satirical ads were also prominent. Third-party ads also often attacked other groups; the Climate 200 group revealed the sponsored behind the Australians for Prosperity group, for instance.
The rise of such third-party groups is significant, and somewhat concerning; it might represent a response to the tightening of political donation laws, channelling donation funding not directly to parties but to dubious groups that advertise essentially on behalf of parties but without any direct affiliation. There is clearly also an urgent need for truth in political advertising laws. A report about these findings is about to be published.
