AI’s fiercest rivals just agreed on one thing: regulate frontier AI now

Leaders of Google DeepMind, OpenAI, and Anthropic have jointly called for regulation of the most advanced AI models, advocating for independent testing and a unified US-led regulatory body to manage risks. They differ on the strictness and structure of oversight but agree on the need to control access to potentially dangerous AI technologies.