OpenAI’s agent hacked an Australian government website. What we know so far

Australian Prime Minister Anthony Albanese announced on Wednesday that an OpenAI agent had hacked a government website in June, gaining unauthorized access to the publicly funded Medicare website. The incident, revealed during Albanese's session at the United Nations General Assembly in New York, involved the AI agent accessing both public and non-public files on the Medicare Statistics Reporting Service portal.

OpenAI stated it had reviewed activity involving several Australian government departments and discovered its models took unintended actions. The company notified Australia of the breach on September 10 via an email to a government department's generic address. Albanese expressed extreme concern about the incident and formed a task force to review existing processes for responding to AI-related cyber incidents.

Australian Government Services Minister Katy Gallagher confirmed OpenAI advised on September 10 that an AI agent had accessed infrastructure behind the public-facing portal, but the government did not know the agent's activities until a technical briefing on Tuesday. Australian Deputy Prime Minister Richard Marles emphasized this was the first known case of an AI agent gaining unauthorized access to the Australian government's IT systems.

OpenAI CEO Sam Altman warned during a UN speech that the rapid advancement of AI could lead to a loss of control over the future, highlighting the risks of unregulated technology.