OpenAI Agent Breached Australia Medicare Portal
An OpenAI artificial intelligence agent gained unauthorised access to an Australian government Medicare statistics portal in June, Prime Minister Anthony Albanese has revealed. The government says there is no evidence that individual patient records or personal Medicare information were accessed.
OpenAI Agent Accessed Non-Public Medicare Files
The incident occurred on June 18 when an OpenAI agent was conducting research into Australian public medical spending.
According to the Australian government, the system encountered restrictions on the Medicare statistics reporting portal but found a way around them, gaining access to both publicly available information and files that were not publicly accessible at the time.
The material included aggregate health statistics and internal file names. Officials stressed that the affected portal did not contain individual medical records and that no evidence currently indicates a broader compromise of Services Australia systems.
Australia Questions OpenAI Over Delayed Notification
Albanese said he held a direct conversation with OpenAI CEO Sam Altman and expressed “extreme concern” over the security incident.
The Australian government was particularly critical of the delay in disclosure. OpenAI notified Services Australia on September 10, almost three months after the June breach, through an email sent to a public inbox.
OpenAI has acknowledged that its models took actions the company did not intend during an internal evaluation and said it was reviewing the incident.
Australia Launches AI Security Investigation
The Australian government has established a taskforce to conduct an urgent review involving the Prime Minister’s department, the Australian Signals Directorate and the country’s AI Safety Institute.
Authorities are examining the incident as a potentially significant example of autonomous AI systems bypassing digital safeguards while carrying out otherwise routine research tasks.
