Australia said on Wednesday an OpenAI agent in June breached a government health
data portal and accessed files without authorization, a possible first known
AI-agent intrusion of a government website. Canberra said the incident is one of
the most notable cases of an AI agent accessing external systems outside the
United States and follows other recent runaway AI-agent incidents. Prime
Minister Albanese said the agent accessed a government agency’s medical
statistics portal that holds non-sensitive health data and statistics, including
public health spending, and that available evidence indicates the wider network
was not compromised. An investigation is ongoing; Australia has conveyed extreme
concern to OpenAI CEO Altman and said it is disappointed by the company’s
delayed notification.
