Australia implements AI accountability requirements for public agencies

The Australian Commonwealth is shifting its approach to artificial intelligence from general governance to specific individual accountability. Under the newly launched APS AI Plan, Public Service Minister Katy Gallagher stated the plan ensures human oversight remains central by requiring a designated accountable owner for every in-scope AI use case by the end of 2026.

While many agencies have appointed Chief AI Officers to manage adoption and capability, recent audits suggest governance structures alone are insufficient. An examination by the Australian National Audit Office of the Australian Taxation Office in 2025 revealed weaknesses in accountability and the integration of governance into project lifecycles, noting that 74 per cent of the agency's in-production models lacked completed data ethics assessments.

This policy shift highlights a broader tension in the AI era regarding the role of human judgment. As intelligent systems become more capable of generating answers and decisions, there is an increasing need for humans to focus on asking the right questions and judging whether specific decisions should be made at all.