Australia's proposed AI safety bill would force frontier model makers to open up to independent testing

Why it matters

Australia's government, led by Prime Minister Anthony Albanese, has proposed a law that would, for the first time, impose mandatory safety rules on companies building frontier AI models in the country. The requirements would apply to the most advanced, highest-risk systems and include independent audits, continuous safety checks, published safety frameworks, and immediate reporting of security incidents to authorities. That reporting duty is partly intended to catch early warning signs from AI agents behaving unpredictably. A new oversight body would enforce compliance alongside the Australian AI Safety Institute, with significant penalties for violations. Deputy Prime Minister Richard Marles and ministers Tony Burke and Tim Ayres framed the move as establishing domestic guardrails before international negotiations begin. The government, working under its agenda for AI in the national interest, says the rules aim to protect citizens while creating the trust and regulatory stability needed to attract investment and let AI benefit the economy and healthcare. Further details will come after consultations conclude.