Anthropic proposes federal AI law letting Washington block risky models
The draft would let courts authorize blocking AI models that pose a sudden peril of mass casualties or over $1 billion in damages. Critics argue its compute threshold is outdated and could tilt the field toward established firms.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
Anthropic has put forward a draft AI safety bill to US lawmakers that would let Washington halt an AI model's release or use if it poses a serious threat, subject to judicial approval. Critics say the bill's frontier compute line of 10^25 FLOPs is outdated and could favor incumbents over startups. Anthropic co-founder Jack Clark has also said an internal model projects AI could lift US economic output to $44 trillion by 2030, and he urged international cooperation on rules for the technology.
Why it matters
The proposal would give the government power to block AI models before harm occurs, and critics say the compute threshold that defines which models fall under it could shape competition in the industry.
Timeline
- Oct 17, 3:01 AM ET: Anthropic presents a draft bill to US lawmakers that would let Washington block an AI model's release or use if it poses an imminent risk of mass casualties or more than $1 billion in damages.
- Oct 17, 12:16 PM ET: Critics say the bill's 10^25 FLOPs frontier line is outdated and could favor incumbents over startups.
- Oct 17, 10:35 PM ET: Anthropic, which is currently under US government sanctions, formally proposes the draft frontier AI safety legislation to federal lawmakers.
- Oct 18, 3:24 AM ET: Clark says an Anthropic internal economic model projects AI could raise US gross output to $44 trillion by 2030 from just over $30 trillion.
Coverage
- Anthropic Proposes 'AI Safety Act' to Congress Amid Government Sanctions (BigGo Finance)
- Anthropic pitches own AI Safety Bill (Times of India)
- Anthropic draft bill proposes AI model shutdowns for major risks (cryptobriefing.com)
- Anthropic’s AI safety bill has an alarming threshold for government action (Mother Jones)
- Anthropic’s Reported AI Regulation Proposal Explained (NeoTeo)
- Anthropic's AI safety bill draws fire over its frontier compute line (Diego Almada Lopez)
- Anthropic proposes law regulating AI to US lawmakers (Dawn)
- Anthropic Proposes Bill to Limit Government Bans on Advanced AI Models (benzinga.com)
- Anthropic's confidential bill limits federal power to halt models (interestingengineering.com)
- Anthropic Pitches Congress a $1B AI Safety Bill (shattered.io)
- Anthropic Quietly Pitches Its Own AI Safety Bill To Lawmakers: Reports (NDTV)
- Anthropic proposes AI safety bill with strict intervention thresholds (Investing.com)
- Anthropic quietly pitches its own AI safety bill to lawmakers, Politico reports (Yahoo News)
- Anthropic Quietly Pitches Its Own AI Safety Bill to Lawmakers, Politico Reports (Reuters)
- Anthropic quietly pitches its own AI safety bill to lawmakers (Politico)
- Anthropic's draft AI bill would restrict shutdowns to extreme cases
- Anthropic pitches lawmakers AI safety bill with model-blocking powers
- Anthropic pitches Congress an AI safety bill as co-founder urges rules
More from Technology & AI
- Eight states sue OpenAI, alleging misleading AI safety claims
- OpenAI announces solution to Navier-Stokes math problem
- Datadog bets AI agents can handle testing of AI-written code
- Anthropic and OpenAI sketch wealth-sharing plans before IPOs
- Mamdani administration proposes AI customer service rules
- Florida engineers study how AI behaves in high-risk situations