TL;DR: Anthropic is backing an AI safety bill that could block model releases over mass-casualty risks, while adding a judicial check on federal intervention.
Summary: Anthropic is pitching an AI safety bill that would let regulators block a model's release if it could plausibly kill 50 or more people. The proposal also reportedly requires federal agencies to obtain a judge's permission before blocking a model, a provision that could limit emergency intervention. The bill is a policy signal for how frontier labs are shaping AI risk legislation.
Why it matters: AI builders should track this bill because it could set a precedent for release gating and judicial oversight of model deployments. Watch for the bill's definitions of risk thresholds and whether the judicial-permission clause survives committee review.
Source: rss