“Anthropic has proposed the creation of a federal agency to regulate AI, which would control which entities are permitted to operate frontier models and would be responsible for conducting safety checks.”