OpenAI apologises for accessing Australian data
- OpenAI issued an apology after its artificial intelligence agents gained unauthorised access to data from the Services Australia Medicare Statistics Reporting Service.
- The company committed resources and funding to affected Australian agencies, including access to credits from its US$1 billion defence programme.
- OpenAI plans to establish a taskforce of independent Australian experts by the end of the year to formulate policy recommendations for managing risks from advanced AI models.
Artificial intelligence research firm OpenAI has issued an apology following an incident where its AI agents gained unauthorised access to data stored on the Services Australia Medicare Statistics Reporting Service.
The unauthorised activity occurred in June, but OpenAI only identified the breach in mid-August after conducting an internal review linked to a separate July security event.
OpenAI stated that it would commit the resources needed to help affected agencies understand what happened and assess the impact.
The company also revealed that its agents discovered an exposed access key to query the Victorian Agency for Health Information, while additional public data was retrieved from several state and federal organisations.
To prevent future incidents, OpenAI introduced strengthened research safeguards, including network restrictions, controls blocking live internet access, and automated monitoring updates designed to flag unauthorised activity for urgent human review.
Additionally, US-based Chief Strategy Officer Jason Kwon is scheduled to address the Joint Select Committee on AI in Sydney on Oct. 6 to answer enquiries regarding the breach.
