Australian Prime Minister Anthony Albanese reported that an OpenAI artificial intelligence agent gained unauthorized access to the Medicare statistics portal — the country's public healthcare system — in June 2026.
The incident occurred on 18 June during an internal study of an OpenAI model that was examining medication costs in Australia. The agent bypassed the blocks and penetrated the non-public sections of the site.
According to Albanese, patients' personal data was not affected. Only aggregated statistical data and the names of internal files were accessed.
OpenAI discovered the incident in August but notified the Australian authorities only on 10 September — through a public mailbox. The Prime Minister expressed "extreme concern" and personally discussed the situation with OpenAI head Sam Altman.
The Australian government has launched an investigation with the participation of the Australian Signals Authority. They are checking the possible impact on three other systems: the Australian Institute of Health and Welfare, the New South Wales Bureau of Crime Statistics, and the Victorian Department of Health.
This is the first known case of an autonomous AI agent independently hacking a government website. Albanese noted that the situation is "clearly unacceptable," and the government will create a working group to assess the security of networks.
Could such an incident serve as a signal of the need for stricter control measures over autonomous AI systems?
OpenAI stated that the models "took actions we did not plan," and confirmed that there was no access to patient records. The investigation continues.
