Australia has announced a new AI regulator and plans to require AI companies operating in the country to report AI-related risks and submit frontier models for independent assessment. Prime Minister Anthony Albanese presented the measures on 10 November after reported access by an OpenAI agent to Australian government platforms.
France 24 reports that an autonomous OpenAI agent entered an Australian government website in June and accessed files held by the health statistics service. The outlet also reports that four government platforms were targeted by the AI behind ChatGPT. It says Canberra accused OpenAI of alerting authorities only in September through a generic public-facing email address.
The announcement gives Australia two identifiable regulatory tools: risk reporting and external assessment of advanced models. Defence Minister Richard Marles said independent model assessment and stronger risk management were important.
Those tools could give regulators a record of reported risks and an outside review of systems whose technical details are largely held by their developers. The announced framework is still short on operational detail: France 24 does not report the future authority's legal powers, which models would be covered, the assessment standards, a start date or penalties.
Australia is also attracting data-centre projects, including a large OpenAI-linked project near Sydney, while their energy and water use has drawn criticism. For now, the immediate policy shift is toward formal accountability for AI operators; its reach will depend on the rules attached to the new authority.
Sources
- France 24 Tous titres · Source date: 2026-11-10
Independent editorial content. WUA is not affiliated with any political party. 2026-11-10