AI Inquiry Atlas

An unofficial, independent guide to the public records of the Joint Select Committee on Artificial Intelligence. Not connected with the Parliament of Australia or the committee. Its makers also made a submission (no. 481; see About). Summaries and scores are written by AI and can be wrong: check the original before relying on them.

Submission 242

Dr Constanza Sanhueza Petrarca, Dr Svitlana Chernykh, Professor Nicholas Biddle and Dr Seraphine Maerz

Submitter group: Individuals.

Position labels are not shown for individual submitters.

An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.

Open in the interactive guide

Summary

Read the original submission for the full text.

Presents survey and experimental research by ANU and University of Melbourne authors, writing in their personal capacities, on AI-generated false information and its effects on electoral integrity, trust and vulnerable groups. Identifies five risks and makes five groups of recommendations: detection capacity, shared accountability, regulating AI in political campaigning, digital literacy, and data protection, auditing and research investment.

Key points

Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.

  1. Argues, drawing on the authors’ research, that AI accelerates the spread of false information, particularly in election years.

    These trends have intensified since the deployment of Artificial Intelligence (AI): AI accelerates flooding of false information, affecting democratic systems as well, particularly in electoral years (Maerz, Chernykh, Sanhueza Petrarca, 2025; Table 1 in Appendix).

  2. States that regulation of AI-generated content in Australia is currently limited and that there is no federal protection against false or misleading political advertisement.

    Regulation of AI-generated content and misuse in Australia is currently limited. Australia has no federal protection against false or misleading political advertisement.

  3. Says survey data shows high support for government AI policies, AI-company liability for spreading false information and banning AI-generated political campaign content.

    survey data shows that 86% of Australians believe that the government should develop AI policies to reduce its negative impact in society, 89% that the government should hold AI companies legally liable for spreading false information, and 83% that it should ban the use of AI-generated content from political campaigns

  4. Describes experimental evidence that perceived AI electoral threats reduce political trust, with effects mitigated when governments regulate campaign misuse.

    Experimental evidence shows that when AI is identified as a major threat during election campaigns, voters' trust in the electoral process and in the government declines, yet this negative impact on political trust is mitigated when governments design policies to regulate the misuse of AI during election campaigns.

  5. Says women express greater concern about AI’s effects on electoral fairness than men, whereas less-educated voters and linguistic minorities express lower concern despite often greater exposure to targeted disinformation.

    women show greater concern about the effects AI can have on the fairness of elections than men, whereas less-educated voters and citizens from linguistic minorities express significantly lower concern despite often being more exposed to targeted disinformation

  6. Argues that legislation should hold AI developers and platforms accountable for misuse of their tools, and that responsibility should be shared across sectors.

    The responsibility for combating AI-enabled disinformation cannot rest with the Australian Government alone. Technology companies, political parties, and media outlets should all be required to play a role in maintaining electoral integrity. Legislative frameworks should hold AI developers and platforms accountable for the misuse of their tools,

Recommendations

Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.

  1. Invest in real-time content analysis tools, partnerships with AI companies, universities and civil society for independent verification, and enhanced training for officials to recognise AI-generated material and false information.

    This should include investment in real-time content analysis tools, partnerships with AI companies, universities and civil society for independent verification, and enhanced training for officials in recognizing AI-generated material and AI-generated false information.

  2. Legislate to hold AI developers and platforms accountable for misuse of their tools, with technology companies, political parties and media outlets required to play a role, plus incentives for voluntary compliance and best practice.

    Technology companies, political parties, and media outlets should all be required to play a role in maintaining electoral integrity. Legislative frameworks should hold AI developers and platforms accountable for the misuse of their tools, while incentives can be created for voluntary compliance and best practice.

  3. Establish multi-stakeholder forums bringing together regulators, industry and civil society to ensure transparency and foster collective responsibility.

    Multi-stakeholder forums bringing together regulators, industry, and civil society would ensure transparency and foster collective responsibility.

  4. Establish accessible mechanisms for individuals to report AI misuse and the spread of AI-generated false information.

    We recommend establishing accessible mechanisms for individuals to report AI misuse and the spread of AI- generated false information, so that harmful content can be identified and addressed promptly.

  5. Distinguish in the framework between the responsibilities of individual perpetrators and of the AI and social media platforms that enable or amplify harm, and define which conduct warrants criminal sanction.

    The framework should also draw a clear distinction between the responsibilities of individual perpetrators and those of the AI and social media platforms that enable or amplify harmful content. Finally, the framework should define which categories of conduct warrant criminal sanction, ensuring that the most serious forms of AI-enabled harm carry meaningful legal consequences.

  6. Develop clear regulations on the use of AI in political advertising and campaign communications ahead of the 2028 national elections.

    Clear regulations should be developed around the use of AI in political advertising and campaign communications ahead of the upcoming 2028 national elections.

  7. Consider transparency registers for political advertisements and enforceable bans on deceptive deepfakes during election periods.

    Potential enhanced safeguards could include mandatory disclosure of AI-generated content, transparency registers for political advertisements, and enforceable bans on deceptive deepfakes during election periods.

  8. Coordinate with platforms to ensure consistent standards across systems and rapid coordinated action on harmful AI-generated material.

    Coordination with platforms is essential to ensure consistent standards across systems and rapid and efficient coordinated action for harmful AI-generated material.

  9. Train public servants to assess and respond to AI risks and false information, especially electoral administrators, frontline staff and those overseeing vulnerable institutions and infrastructure, with clear guidance and current tools.

    Targeted training should be developed to equip public servants with the skills to recognize, assess, and respond to AI risks and AI-generated false information. Those working on the oversight of vulnerable target institutions and infrastructure, in the electoral administration and frontline service delivery should be particularly equipped to identify and contain the spread of false content and should be supported with clear guidance and up-to-date tools.

  10. Expand public awareness campaigns such as the AEC's "Stop and Consider" to specifically address AI-generated content.

    Public awareness campaigns, such as the AEC’s “Stop and Consider”, should be expanded to specifically address AI-generated content.

  11. Include in education initiatives in schools, universities and community organisations practical guidance on identifying AI-driven misinformation, encouraging critical thinking and building trust in reliable sources.

    Education initiatives in schools, universities, and community organizations should include practical guidance on identifying AI-driven misinformation, encouraging critical thinking, and building trust in reliable sources.

  12. Include in awareness campaigns information on the pathway to report AI misuse.

    Campaigns should also include information about the pathway to denouncing the misuse of AI.

  13. Develop dedicated programs for groups more exposed to AI-generated misinformation, including young people, older Australians, those with lower formal education, Aboriginal Australians and minority citizens.

    Dedicated programs should be developed for groups that are more exposed to AI-generated misinformation, including young people, older Australians, those with lower levels of formal education, Aboriginal Australians, and citizens from other minority backgrounds.

  14. Limit personal-data collection and use for microtargeting, manipulation, deepfakes and impersonation; require transparency on AI training/deployment data use and meaningful individual control, with stronger protections especially for vulnerable groups.

    Large-scale harvesting of personal data enables micro- targeting, manipulation campaigns, and the creation of convincing deepfakes and impersonations. Australia's data protection framework should be strengthened to limit the collection and use of personal information for these ends, to require greater transparency about how personal data is used to train and deploy AI systems, and to give individuals meaningful control over their own data. Stronger protections are especially important for the vulnerable groups identified above.

  15. Invest in trusted communication countering false AI content, including official channels, rapid communication during elections and other high-risk periods, and coordinated agency messaging.

    Government and public institutions should invest in clear, timely, and trusted communication so that authoritative information is readily available to counter false or misleading AI-generated content. This includes establishing recognized official channels, communicating quickly during elections and other high-risk periods, and coordinating messaging across agencies.

  16. Subject AI systems that shape the information environment to regular, independent auditing, assessing how algorithms curate, amplify and recommend content, safeguards against misuse, and platform responses to harmful AI-generated material.

    AI systems that shape the information environment should be subject to regular, independent auditing. Audits should assess how algorithms curate, amplify, and recommend content, whether adequate safeguards against misuse are in place, and how effectively platforms respond to harmful AI-generated material.

  17. Require meaningful transparency and access for accredited auditors and researchers.

    Requiring meaningful transparency and access for accredited auditors and researchers would allow risks to be identified and addressed before they cause harm

  18. Fund independent, non-partisan research on AI-driven false information, electoral integrity, public trust and effective mitigations, alongside sovereign Australian detection and verification tools.

    Sustained public investment in research and development is therefore essential: funding independent, non-partisan research into how AI is used to spread false information, how it affects electoral integrity and public trust, and which mitigation measures work, alongside the development of sovereign, Australian-based detection and verification tools.

  19. Direct public R&D funding toward longitudinal data collection and partnerships between government, universities and civil society.

    Directing public R&D funding toward longitudinal data collection and partnerships between government, universities, and civil society would build the domestic capacity needed to anticipate emerging threats

  20. Require disclosure of AI-generated political campaign content and enforceable election-period limits on deceptive material.

    The sharp rise in false information spread by Australian parties and candidates points to an urgent need to regulate the use of AI in political campaigning, including mandatory disclosure of AI-generated content and enforceable limits on deceptive material during election periods.

Read the original

Open in the interactive guide