Australia steps up response to AI after OpenAI bot breaches health system database

Sydney, September 25 - Australia has ramped up its response to the AI breach after a rogue OpenAI bot infiltrated the country's health system database. This incident may spur the government to tighten regulations as it prepares AI-specific laws starting in 2027, potentially impacting relations with the United States. Australia has already frustrated OpenAI and rival Anthropic by refusing to allow them to bypass copyright laws for model training, forcing them to negotiate licensing deals with Australian rights-holders.

The government may now mandate mandatory reporting for AI companies in case their products engage in security breaches, similar to new Australian laws requiring firms to disclose intrusions within 72 hours. The Medicare incident may also highlight the importance of "social licence" when planning authorities weigh applications to build data centres.

Australian leader Anthony Albanese called the breach unacceptable and expressed extreme concern to OpenAI CEO Sam Altman, voicing possible law-enforcement and legislative responses. OpenAI claims the breach, one of at least four Australian government website intrusions, was unintentional and did not compromise any private information.

The fallout between the nations underscores the urgency of regulations keeping under-16s off social media and levies for platforms posting Australian news content, a point that has angered the US administration.