Australia Demands More AI Safeguards After Revealing OpenAI Hack

(Bloomberg) -- Australian Prime Minister Anthony Albanese used his trip to the United Nations to call for greater controls on artificial intelligence, pointing to the revelation that OpenAI had hacked a government website as evidence of the need for tougher safeguards.

"Recently, an artificial intelligence agent infiltrated an Australian government website. This is unacceptable and it is why Australia has been leading the way in establishing Australian standards for AI," Albanese said in New York, adding that countries needed to work together to shape how the technology develops.

Listen and follow The Bloomberg Australia Podcast on Apple, Spotify, YouTube or wherever you get your podcasts.

The minister in charge of AI policy, Andrew Charlton, told national broadcaster ABC that the OpenAI breach "makes the case" for the government's safety goals, adding that he aimed to pass new laws early next year on AI safety and data center construction. The measures would seek to manage risks highlighted by the June hack of a government website, he said.

While the intrusion into a government health-statistics website was minor, the incident underscored the risk that AI models can circumvent security barriers without human control. Albanese and other ministers have also criticized how long OpenAI took to notify Canberra, with almost three months elapsing between the breach and the company sending notice to a generic government email address.