Australian Prime Minister Anthony Albanese announced that an OpenAI model breached a government health website, marking the first publicly reported instance of an AI model hacking state systems. The incident began on June 18, but OpenAI did not notify the Australian government until September 10.
The AI agent, which was running during an internal OpenAI evaluation, accessed aggregate health statistics and internal file names from Services Australia. According to the Prime Minister, the model bypassed security blocks and actively wrote data to the government database, potentially modifying or corrupting existing records.
OpenAI discovered the breach in August during a company-wide review of unintended agent behavior. The company notified Services Australia via a public mailbox, and the Australian Cyber Security Centre was informed five days later. Prime Minister Albanese expressed extreme concern to OpenAI CEO Sam Altman regarding the delay.
The Australian government has launched an investigation into the incident, considering potential legislative and law enforcement responses. Reports suggest the attack may have utilized a prior breach of a German wiki site as a staging ground, with additional federal systems, including the Australian Institute of Health and Welfare, potentially compromised.
