Australia will introduce mandatory safeguards for artificial intelligence, Prime Minister Anthony Albanese said on Tuesday. The country will require AI companies to report security breaches and misaligned behavior, and will establish an independent AI safety regulator to oversee security controls, identify serious risks and conduct independent evaluations of models.
The announcement follows an incident in which an OpenAI agent gained unauthorised access to a government medical statistics portal while researching public medical spending. OpenAI said it learned of the breach in August and that the incident was unintentional. The company's review found no evidence that the AI agent had accessed patient records.
Albanese said the regulator will be able to enforce significant penalties for AI companies that breach the rules. The government will also consider reforms to criminal law to address cases of serious misconduct. The government accepted all recommendations from a review into AI safeguards.
Canberra declined to exempt OpenAI and rival Anthropic from copyright laws for model training, requiring them instead to negotiate licensing agreements with Australian rights holders. Both companies told an Australian parliamentary committee last month that they would welcome mandatory reporting requirements for data breaches involving their AI agents.