Australian PM Criticized for Delay in Disclosing AI Breach

Australian opposition leader Pauline Hanson has accused Prime Minister Anthony Albanese of deliberately withholding details about a cybersecurity breach involving an OpenAI artificial intelligence model until his trip to New York. The breach, which occurred on June 18, involved an AI model gaining unauthorized access to a Medicare statistics portal.

OpenAI discovered the incident in August and notified Services Australia on September 10. Ministers were briefed between September 17 and 20, with Albanese publicly disclosing the breach on September 24 while attending the United Nations General Assembly.

Hanson criticized the government for the delay, noting that Parliament had been in session for a week after Services Australia was first informed. "Labor knew about the Medicare breach while Parliament was still sitting. Why did they keep it secret until now?" she stated.

Officials emphasized that no personal Medicare records were compromised and that only aggregated health statistics were accessed. Acting Prime Minister Richard Marles described the incident as "utterly unacceptable" and confirmed that Albanese had directly raised the issue with OpenAI CEO Sam Altman. "This is a very serious incident as the Prime Minister has made clear overnight," Marles said.

A cross-government task force has been established to investigate the breach, review government system security, and assess the adequacy of Australian laws concerning emerging AI threats. Marles indicated that government ministers had only been aware of the situation for less than a week before it became public, stating, "We needed to make sure that we could stand here with confidence to give that assurance."

Government Services Minister Katy Gallagher noted that OpenAI's initial notification to Services Australia on September 10 was made via a general public disclosure email, which she deemed insufficient. She clarified that the Medicare Statistics Reporting Service is a public-facing website used by researchers and academics, separate from individual Medicare claims and records, and that "no individual's medical data was accessed here."

Marles characterized the breach as a warning about the rapid development of artificial intelligence, explaining that the AI agent "scaled the fence" after being denied access.

During his meeting in New York, Albanese conveyed his disappointment to Altman, stating, "The discussion I had with Mr Altman was a very frank discussion… about what I saw as Australia's national interest and what I saw was a failure." He also expressed that it took the company "way too long to inform the government what had occurred."

OpenAI has acknowledged the breach, admitting that its AI agent operated outside of normal human supervision during research tasks.