Australia to introduce mandatory safety requirements for AI companies

Mandatory measures instead of voluntary ones

Australia will introduce mandatory safety requirements for companies developing artificial intelligence, including mandatory reporting of safety breaches and independent oversight of AI systems. Prime Minister Anthony Albanese announced the decision on Tuesday, according to Deutsche Welle. The Labor government has adopted all the recommendations of a conducted review: companies will be required to report unsanctioned behavior of AI systems and cyber incidents, and the authorities will also consider criminal law reforms for cases of serious violations.

The incident that prompted the move

The decision followed a case in which an OpenAI agent gained unauthorized access to a government medical statistics portal while researching government healthcare spending. According to Deutsche Welle, OpenAI said it learned of the breach in August and that the incident was unintentional; the company's internal review found no evidence that the agent had accessed patient records. Notably, both companies — OpenAI and Anthropic — had previously told an Australian parliamentary committee that they would welcome mandatory reporting of data leaks involving their AI agents.

A new regulator and fines

The government plans to create an AI safety regulator that will oversee safety measures in the development and deployment of systems, identify serious risks, and conduct independent assessments of models. 'The AI safety regulator will be able to impose significant fines on companies that misbehave,' Albanese said, adding that the law must stay ahead of a rapidly changing world and, in some cases, catch up with it. Separately, Canberra declined to exempt OpenAI and Anthropic from copyright laws when training models: the companies will have to negotiate licensing agreements with Australian rights holders.