From Mother Jones · Sophie Hurwitz · · 1m
Anthropic's AI Bill Sets High Risk Threshold
Anthropic's proposed AI safety bill requires government action only for risks causing over $1 billion in damage or 50+ deaths.
In brief
Anthropic's proposed AI safety bill requires government action only for risks causing over $1 billion in damage or 50+ deaths. Anthropic's draft bill lets the government block AI models only if they pose a foreseeable risk of over $1 billion in damages or killing 50+ people. Originally reported by Mother Jones.
Anthropic's Capitol Hill Lobbying
Anthropic has ramped up lobbying efforts in the first half of 2026, spending more than it did in all of 2025.
Bill's High Risk Threshold
The bill would let the government block an ai model only if it poses a foreseeable risk of more than $1 billion in damages or killing 50+ people.
Limited Risk Categories
The bill defines intervention-worthy risks as only creating biological weapons, evading developer controls, or cyberattacks.
State-Level Lobbying
Anthropic spent at least $90,000 in california to push pro-data-center policies and deployed lobbyists in tennessee and north carolina.
The takeaway
Anthropic's draft bill lets the government block AI models only if they pose a foreseeable risk of over $1 billion in damages or killing 50+ people.
