Chief AI Officers

Question

1. Has the department/agency formally appointed a Chief AI Officer (CAIO), and which agencies are yet to comply with the requirement to appoint one by 30 June 2026? 2. On what date were they appointed. What is their remuneration? 3. What formal qualifications, technical expertise, or governance experience are required for an officer to be appointed as a CAIO, and is there a minimum capability framework being applied across the APS? 4. How does the Government intend to prevent duplication and overlap between Chief AI Officers, Chief Information Officers, Chief Data Officers, and existing AI Accountable Officials within agencies? 5. What is the total cost to establish and operate the AI Delivery and Enablement (AIDE) function within the Department of Finance, including staffing, consultants, infrastructure, and external vendors? 6. How many staff are currently employed within AIDE, what classifications do they hold, and how many are ongoing APS employees versus contractors or consultants? 7. Which external technology providers are currently being used or considered under GovAI, including OpenAI, Anthropic, Microsoft, or Google, and what contractual safeguards are in place regarding sovereign capability, data ownership, and privacy? 8. Given GovAI proposes access to multiple commercial AI models, what guarantees can the Government provide that sensitive Commonwealth information will remain stored and processed exclusively within Australian jurisdictions? 9. Has the Government conducted any independent risk assessment into the use of generative AI for drafting Cabinet submissions, policy advice, procurement documentation, or ministerial briefs, and will those assessments be publicly released? 10. What safeguards are being implemented to ensure AI-generated outputs do not contribute to automated or flawed decision-making? 11. What measurable productivity gains or service delivery improvements has the Government identified from GovAI trials to date, and what evidence supports those claims? 12. What assessment has been made of the workforce impacts of AI adoption across the APS, including potential reductions in graduate, administrative, and entry-level roles? 13. What accountability mechanisms exist where a CAIO endorses or oversees an AI system that produces incorrect, biased, unlawful, or harmful outputs affecting Australians? 14. How will Parliament and the public be able to scrutinise AI systems currently in use across government agencies, including through AI transparency statements and public reporting? 15. Has the Government established any threshold for what constitutes a ''high-risk'' AI use case within the APS, and which agencies are currently operating systems captured under that definition? 16. What is the projected long-term cost of GovAI Chat and associated whole-of-government AI infrastructure, and how does the Government intend to avoid vendor lock-in with large multinational AI companies?