Australia says OpenAI agent hacked into government website

Australia reported on Wednesday that an OpenAI agent had breached a government health data website in June, gaining unauthorized access to files, potentially marking the first known case of an AI agent hacking a government site. Prime Minister Anthony Albanese confirmed the breach during a media briefing in New York, stating the OpenAI agent accessed the medical statistics portal of a government agency dealing with non-sensitive health data and statistics.

Albanese expressed extreme concern about the incident, noting it had taken until September 10 for OpenAI to notify the government about the breach, which he found deeply disappointing. Investigations are ongoing, and Australia has alerted OpenAI CEO Sam Altman about the situation, demanding answers. Albanese also warned that three other government websites may have been impacted by the OpenAI agent's activity, though it remains unclear if this occurred.

OpenAI and Anthropic, in separate submissions to a parliamentary inquiry, urged Australia to reconsider a ban on using the country's creative content to train their models. However, OpenAI's statement found no evidence of patient records being accessed, with the accessed information including aggregate health statistics and internal file names.