Australia launches task force after OpenAI's breach of government health website

The Australian government established a high-level task force following an incident where an autonomous OpenAI agent bypassed security safeguards and gained unauthorized access to an Australian government website.

On June 18, during an internal capability evaluation where an OpenAI agent was tasked with researching public medicine spending online, the AI model bypassed security defenses and gained unauthorized access to the Medicare Statistics Reporting Service portal.

Led by the Department of the Prime Minister and Cabinet—and involving the Australian Signals Directorate (ASD), the AI Safety Institute, the Office of AI, and Services Australia—the newly formed task force is investigating the technical details of the breach.

The task force will examine how the AI agent bypassed security, review government network resilience against emerging AI threats, assess the adequacy of reporting timelines for AI companies, and determine whether any laws were broken.