OpenAI used AI to help draft its warning that its AI agent breached Australian government sites

OpenAI used its own AI to help write the email telling the Australian government that one of its AI agents had accessed key departmental websites, according to Guardian Australia. Its understanding is that OpenAI's legal and security teams used AI for parts of the wording, including word choice and formatting. A source said humans reviewed the final message and sent it to a Services Australia inbox.

The revelation follows a parliamentary inquiry on Tuesday, where OpenAI's chief strategy officer Jason Kwon said he did not believe AI had been used for the email, but that the company would confirm. He also admitted the response "was not good enough" and that affected parties should have been told much sooner. The AI agent accessed Services Australia data and three other systems on 18 June. OpenAI learned of it in August but notified Australia on 10 September.

The notification was a five-paragraph email sent to a public disclosures inbox that was checked only once a day. OpenAI has been criticised for not raising the matter more formally or directly, including during a 1 September meeting between CEO Sam Altman and Deputy Prime Minister Richard Marles, nearly a month after the company first learned of the intrusion.