All Documents Relied On Or Relevant To The Attorney General’s Media Statements On 2 And 3 September Relating To The Involvement Of AI In Making FOI Applications.

Written by Claude Haiku 4.5 from the extracted text on 1 Mar 2026. It can miss or misread redacted and scanned pages. Check the source before relying on it.

The Attorney-General's Department released internal communications and talking points (dated 30 August 2025) prepared for media statements on 2-3 September 2025 regarding artificial intelligence's role in Freedom of Information applications. The documents include email correspondence, draft talking points marked as Cabinet-in-confidence, and supporting materials discussing government agency concerns about high-volume, AI-generated FOI requests. Key disclosed positions include: consultation with the National Intelligence Community about foreign adversaries potentially exploiting the FOI scheme to obtain innocuous information for intelligence purposes; domestic examples of AI generating both FOI requests and draft responses, with one agency reporting approximately 20% of personal and 50% of non-personal FOI matters containing AI-generated content; and international precedents, including US election workers receiving thousands of AI-generated FOI requests during the 2020 presidential election. The talking points indicate government consideration of requiring named requesters rather than anonymous submissions to deter potential misuse. Significant portions remain redacted under section 22(1) (legal professional privilege) and section 47E(d) (Commonwealth security), reflecting the classified nature of National Intelligence Community advice.