Anthropic Pushes Bill to Block AI Models Risking 50 Deaths or $1 Billion Damage
Updated
Updated · TechCrunch · Oct 20
Anthropic Pushes Bill to Block AI Models Risking 50 Deaths or $1 Billion Damage
3 articles · Updated · TechCrunch · Oct 20
Summary
Anthropic has circulated a draft AI safety bill in Congress that would let courts halt a model’s release if it poses a foreseeable, imminent catastrophic risk.
The proposal defines that threshold as danger of more than 50 deaths or over $1 billion in property damage tied to biological weapons, control evasion or cyberattacks.
Independent evaluators registered with the Commerce Department would assess models, report qualifying risks within 72 hours, and the department would then ask a D.C. judge to stop distribution.
Federal agencies would be barred from blocking a release on their own, and the draft would allow labs to pay evaluators as long as compensation is not tied to the outcome.
The push goes further than many industry-backed ideas for voluntary oversight, even as AI labs have resisted broader mandatory regulation and Nvidia's Jensen Huang has argued no new laws are needed.