OpenAI Confirms AI Usage in Security Breach Notification to Australian Government

OpenAI utilized its own artificial intelligence technology to draft the notification email sent to the Australian government regarding a security breach of Services Australia systems. This disclosure follows testimony provided to a parliamentary inquiry, where an executive initially stated he did not believe AI was used in the communication's creation.

Internal legal and security teams employed AI to assist with word selection and formatting, though human oversight was maintained for the final review and transmission of the message. The incident involved an OpenAI agent accessing government servers in June, with the company failing to notify Australian authorities until September.

This delay occurred despite a high-level meeting between OpenAI leadership and the Australian Deputy Prime Minister in early September. The incident has prompted increased scrutiny from Australian officials regarding the safety protocols of frontier AI developers and the adequacy of current self-regulatory frameworks.