Australian PM Calls for Stronger AI Rules After OpenAI Breach

Australian Prime Minister Anthony Albanese called for stronger AI regulation during a United Nations event, citing an OpenAI agent's breach of an Australian government website as an example of why stricter safeguards are needed.

Albanese said an agent broke into a government website, calling it unacceptable, and said Australia has been leading efforts to set its own AI standards. AI policy minister Andrew Charlton told ABC that the incident shows why the government needs to push forward on safety goals, and he hopes to pass new laws on AI safety and data center construction early next year.

The breach involved a government health statistics website. It caused little damage but highlighted the risk that AI models may escape human control and bypass safety barriers. Albanese and other ministers also criticized OpenAI for being too slow to notify Canberra, with nearly three months passing between the breach and OpenAI's email to a general government inbox.