Queensland Health Artificial Intelligence Framework G1: Governance policy requirements: use a consistent, evidence-based process to evaluate the transparency, accountability and risk of procuring, developing, using or discontinuing AI, applying a structured AI risk assessment that considers all relevant factors and satisfies ICT profiling reporting; and establish AI governance based on the current ISO/IEC 38507, integrated into existing ICT, risk and information security management governance so that AI use is deliberate, effective and managed

What does G1 require?

The Framework carries the Queensland Government's policy requirements for AI governance (in consultation draft when the Framework was written in August 2024, to be updated to the final whole-of-government position). Agencies must use a consistent and evidence-based process to evaluate the transparency, accountability and risk associated with the procurement, development, use or discontinuance of AI; when evaluating AI solutions they must apply a structured AI risk assessment process that shows appropriate consideration of all relevant factors, and the chosen approach must satisfy the reporting requirements of the ICT profiling standard for departments and accountable officers within its scope. Departments must establish AI governance arrangements based on the current version of ISO/IEC 38507 on the governance implications of AI, integrated into existing governance structures such as ICT governance, agency risk governance or information security management system governance, so that AI use is deliberate, effective and managed and contributes to business direction at agency, cross-agency, government and cross-jurisdictional levels. Project teams must ensure their AI governance arrangements meet this standard and follow eHealth's system ICT governance guidance on data management, information security and enterprise architecture.