Anthropic chief calls for legal blocks on dangerous AI and tougher oversight

Anthropic CEO Dario Amodei argues that governments should have the power to legally block or deter dangerous AI deployments, proposing a sweeping regulatory framework that includes mandatory testing and auditing of frontier models, tighter international coordination, and economic safeguards to manage AI-driven disruption; he also urges reforms to agencies and faster pre-release approvals for AI-related breakthroughs.