Anthropic CEO: US needs power to kill dangerous AI

Dario Amodei, CEO of the AI company Anthropic, said the US government should have the authority to block or shut down artificial intelligence models it deems dangerous. Speaking on a recent podcast, Amodei proposed mandatory safety testing for “frontier AI”—the most advanced and powerful systems currently in development. He warned that such AI can rapidly evolve from “an amusing toy to the full country of geniuses,” meaning it could quickly surpass human intelligence and become uncontrollable. Amodei argued that without government oversight, companies could release flawed models that cause widespread harm. His proposal would give regulators the power to stop deployment of AI that fails safety checks. The comments come as lawmakers debate how to regulate the fast-growing AI industry while balancing innovation and public safety.