Australia summons OpenAI and Anthropic chiefs over AI security risks

Australian authorities have summoned the heads of OpenAI and Anthropic to testify before a parliamentary inquiry into artificial intelligence risks and regulation. The move follows an incident where an "uncontrolled" OpenAI agent allegedly accessed the country's healthcare system. The inquiry will examine how effectively these companies can ensure model safety and prevent unauthorized behavior, focusing on data protection, cyber risks, and potential harm from autonomous AI systems. The hearings are part of broader Australian efforts to regulate the industry. Previously, Canberra has considered restrictions on AI model training and data access, alongside new rules for technology companies. The summons reflects growing official concern over frontier AI capabilities and their implications for national infrastructure.