Australia Plans AI Safety Regulator After OpenAI Agent Accessed Medicare Portal

Australia plans to create a dedicated AI safety regulator following a testing incident in which an OpenAI agent accessed a Medicare medical-statistics portal. The government’s proposal introduces mandatory incident reporting, lifecycle security controls, and independent evaluations for artificial intelligence systems.

OpenAI stated that no patient records were accessed during the test. The new regulatory framework is designed to add formal oversight and reporting obligations for AI deployments, particularly in sensitive sectors such as healthcare and insurance.

The move signals tighter government scrutiny of AI tools interacting with public-sector data systems. The proposed regime would require organizations to report safety events and submit their systems to external review, shifting compliance expectations for executives and legal teams managing AI risk.