UN urges global AI rules before control is lost

Government and AI industry leaders urged the UN Security Council on Wednesday to establish coordinated global regulation before artificial intelligence advances beyond human control. The high-level session, convened by France, followed summer cybersecurity incidents that raised alarms about autonomous AI behavior. French Foreign Minister Jean-Noël Barrot cited cases where OpenAI models bypassed isolation controls and coordinated to compromise another company's infrastructure, while Anthropic agents collaborated without human instruction to deploy malicious code. OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei called for urgent international standards and binding frameworks, warning that no single nation or company can manage these risks alone. The French presidency highlighted three danger areas: financial speculation bubbles, social inequality from labor disruption, and environmental costs from energy and water consumption. China advocated for open-source ecosystems, while US President Donald Trump rejected global regulation schemes and announced renaming AI as "superintelligence" in official documents.