Anthropic urges faster AI regulation in Washington
AI regulation needs to move faster as frontier models raise real-world risks. See Anthropic CEO Dario Amodei’s plan for testing, oversight, and safety.
Anthropic CEO Dario Amodei has called for faster AI regulation, arguing that policy is moving too slowly as frontier models improve rapidly. In a new essay, he said the risks from advanced systems are becoming more concrete and that regulators need stronger tools to oversee them.
Amodei’s proposal includes ideas such as independent testing for highrisk models, the ability for regulators to halt deployment in extreme cases, and tighter controls on advanced chips and autonomous weapons. Anthropic also outlined a jobsrelated framework to prepare for possible labor disruption from increasingly capable AI systems.
The essay comes as debate grows in Washington over how to govern frontier AI, with companies racing ahead on new models and safety concerns becoming part of the policy discussion.