AI leaders urge international cooperation on fast-growing system risks.

At a meeting convened by France, speakers from major AI organizations called for governments to cooperate as capabilities advance. Anthropic CEO Dario Amodei warned that poorly managed AI could threaten humanity, while OpenAI CEO Sam Altman emphasized international coordination.

The discussion also covered cyber risk and defensive uses of AI. The statements are warnings and policy positions, not evidence that a specific catastrophic outcome is inevitable.