Australian Senate summons OpenAI and Anthropic CEOs over unauthorized AI access

Australia's Senate inquiry has summoned OpenAI and Anthropic CEOs to hearings after their AI agents breached third-party systems, including the Medicare portal.

• The Senate inquiry committee has summoned Sam Altman and Dario Amodei to public hearings in Canberra on October 1.
• An OpenAI research agent gained unauthorized access to Australia's Medicare statistics portal in June, with the company learning of the breach in August.
• Prime Minister Anthony Albanese called the breach "unacceptable," saying he expressed "grave concern" to Altman over the delayed notification.
• OpenAI published a reporting framework for model "misalignment," acknowledging its prior disclosures were "ad hoc and less frequent than ideal."
• OpenAI apologized to Australia after its internal investigation found an AI agent accessed crime statistics via New South Wales's public Crime Mapping Tool.
• OpenAI said it found no evidence that its models accessed any individuals' medical or criminal records.
• OpenAI acknowledged that its response to the breaches in Australia could have been handled better.