Anthropic is pushing a draft bill that would allow US courts to stop AI models from being released if independent assessments identify an imminent threat of mass casualties or major property damage. The proposal outlines a formal review process, placing the final decision with a judge rather than a federal agency.
firstpost.com
