Leaked Anthropic Draft AI Regulatory Bill Is Unacceptable and Troubling
Zvi (Don't Worry About the Vase) 14 hours ago 30
A leaked draft federal AI regulatory bill said to be shopped by Anthropic has triggered widespread criticism from AI safety commentators for being weaker than prior proposals like the FRONTIER Act. The bill would require an imminent catastrophic-risk bar tied to “more than 50 people” killed or “more than a billion dollars” in damage before federal restrictions could be imposed. If enacted, it would more strongly preempt state AI rules while providing limited enforcement and narrower shutdown authority, which critics say would leave people less protected than existing law.
