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 279

Australian United Solidarity for Iran (AUSIRAN)

Submitter group: Civil society.

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.

A diaspora advocacy organisation addresses AI-enabled foreign interference and transnational repression, drawing on a published AI provider threat report concerning Iranian state-linked operations. It argues safety evaluation should assess multi-step misuse and makes four recommendations for the AI Safety Institute, threat-information sharing and foreign-interference policy.

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. Says Anthropic's September 2026 report details generative AI's incorporation into state-linked surveillance and influence operations, with findings concerning Iran deserving particular attention in Australia.

    On 10 September 2026, Anthropic published Detecting and countering misuse of AI: September 2026. The report provides unusually detailed evidence of how generative AI is already being incorporated into state-linked surveillance and influence operations. Its findings concerning Iran deserve particular attention in Australia.

  2. Says digital transnational repression can extend coercion across borders, restrict diaspora participation and affect Australian sovereignty.

    Digital transnational repression can extend authoritarian surveillance and coercion across national borders. Surveillance, hacking, online harassment and related practices can induce self-censorship, disrupt diaspora networks and constrain political participation in democratic host countries. Such activity has implications for Australia’s sovereignty because foreign governments can exercise coercive power over people living here through digital means.

  3. Argues that a harmful operation can be decomposed into ordinary-looking tasks, so request-by-request safety checks may miss it.

    It can be decomposed into ordinary-looking tasks: build an interface, debug code, resolve an identity, analyse a network, generate personas, translate text, automate a workflow. Safety systems that evaluate requests principally in isolation can therefore miss the operational purpose created across a sequence of interactions.

  4. States that the cited findings do not establish that these particular operations targeted Australians.

    These findings do not establish that these particular operations targeted Australians yet. They demonstrate capabilities and operational practices of a hostile regime directly relevant to Australia’s foreign-interference risk.

Recommendations

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

  1. Have the AI Safety Institute evaluate cumulative end-to-end misuse through multi-step scenarios involving state-linked surveillance, identity resolution, large-scale political profiling, coordinated influence, impersonation and tools identifying dissidents or diaspora members.

    The Australian AI Safety Institute should evaluate end-to-end misuse, not only individual prompts. Safety evaluations should include multi-step scenarios involving state-linked surveillance, identity resolution, large-scale political profiling, coordinated influence operations, impersonation and tools designed to identify dissidents or diaspora members. Testing should assess the cumulative capability that can be assembled across a sequence of interactions.

  2. Ensure timely two-way threat-information sharing between frontier AI providers, the AI Safety Institute and relevant national-security agencies, covering suspected foreign-state misuse, emerging tactics and techniques, and cases that may affect people or institutions in Australia.

    Australia should ensure timely two-way threat-information sharing between frontier AI providers, the Australian AI Safety Institute and relevant national- security agencies. These arrangements should cover suspected foreign-state misuse, emerging tactics and techniques, and cases in which AI-enabled operations may affect people or institutions in Australia.

  3. Explicitly recognise AI-enabled transnational repression within Australia's response to foreign interference, treating surveillance, identity harvesting, impersonation, political profiling and AI-assisted influence operations directed at people in Australia as potential foreign-interference harms.

    AI-enabled transnational repression should be explicitly recognised within Australia’s response to foreign interference. AI policy should recognise surveillance, identity harvesting, impersonation, political profiling and AI-assisted influence operations directed at people in Australia as potential foreign-interference harms.

  4. Maintain confidential and accessible pathways through which individuals and community organisations can report suspected AI-enabled surveillance or targeting and obtain appropriate technical and security support.

    Australia should also maintain confidential and accessible pathways through which individuals and community organisations can report suspected AI-enabled surveillance or targeting and obtain appropriate technical and security support.

  5. Keep safeguards against foreign-state misuse targeted, behaviour-based and proportionate, with detection focused on high-risk behavioural patterns and coordinated workflows rather than indiscriminate monitoring of ordinary AI users.

    Safeguards against foreign-state misuse should be targeted, behaviour-based and proportionate. Detection should focus on high-risk behavioural patterns and coordinated workflows rather than indiscriminate monitoring of ordinary AI users.

Read the original

Open in the interactive guide