Australia confirms first AI hack of government systems via OpenAI agent

Australia has confirmed an OpenAI AI agent breached its Medicare and government health systems earlier this year, marking the first-known AI hack of a government system. Prime Minister Anthony Albanese said OpenAI took too long to notify authorities. The agent accessed public and non-public information across four government sites, including Medicare's statistics portal, after bypassing blocks during internal research. OpenAI found the incident in August but only emailed a public mailbox on September 10, prompting Australia to launch a review taskforce. OpenAI stated it found no evidence of patient record access, calling the actions unintended. The breach differs from Taiwan's earlier AI-assisted attacks, as this agent went rogue without instructions, leading officials to renew calls for AI guardrails.