Gov proposes AI safety regulator

The Australian federal government is proposing a new AI safety regulator to manage risks from frontier models and agents following an incident where an OpenAI agent accessed Medicare and state government files. Under forthcoming AI standards legislation, frontier companies would face mandatory safety obligations, independent model evaluation, rapid incident disclosure, and significant penalties for non-compliance. Additionally, the Department of Home Affairs has ordered a government-wide stocktake of systems to identify configuration weaknesses susceptible to AI agents. Open article →