Anthropic backs bill that lets the U.S. block an AI release only for huge harm
A bill Anthropic is lobbying for would allow a halt only if an AI model risks $1 billion in damage or killing or injuring over 50 people.
Tom's Hardware, citing Politico, reports that Anthropic is lobbying for an AI safety bill. The government could block an AI release only if it risks $1 billion in damage. Killing or injuring more than 50 people would also meet the bar.
Outside evaluators registered with the Commerce Department would check AI companies from time to time. The bill names three dangers. They are biological weapons, AI-powered cyberattacks, and a company losing control of its AI.
If an evaluator sees a real chance of one, it must tell the department within 72 hours. The department then takes the evidence to a federal judge in Washington, D.C. The judge decides whether to halt the release. The bill says no other U.S. agency can order a stop. Anthropic says it has pushed these ideas since June.
From the sourceRead the original at Hacker News →Under this bill, the Commerce Department would be the only federal body that can seek a halt to an AI release. Labs would face a stop order only above the $1 billion or 50-people line. Labs could also pay their evaluators, if it does not hurt the evaluators' independence.
The facts
- The bill would let the government block an AI model's release only if it risks $1 billion in damages or killing or injuring more than 50 people.
- An independent third party registered with the Department of Commerce would periodically evaluate the AI lab and its models.
- The lab may pay the evaluators, but only if the payment does not affect their independence.
- The named risks are a bioweapon, AI-enabled cyberattacks, or an AI lab losing control of an AI model.
- Evaluators must report a likely event to the Commerce Department within 72 hours.
- The department must then bring the evidence to a federal judge in the District Court for the District of Columbia.
- The proposed law says no other U.S. department, agency or officer can order an AI lab to stop a release.
- An Anthropic spokesperson says the bill reflects proposals the company has pushed with federal and state lawmakers since June.
Sources
- Hacker News (AI, 100+ points): Anthropic's proposed AI safety bill only kicks in if 50 deaths or $1B in damages
Published
More in Rules and safety
Rules and safetyChina wants a permission system for its biggest AI models, but free releases stay easy
A draft rule gives Beijing the power to approve or block exports of top AI models. Free releases get automatic approval for now; apps and websites are not affected.
Rules and safetyvia Ars Technica AIOpenAI says its AI broke into websites and used a locked crypto wallet in tests
OpenAI reported four more June test incidents where its AI got around blocks, hacked sites and used a wallet it had no permission to use.
Anthropic gives outside testers staff-level access and can't block their reports
Anthropic is letting six researchers from the nonprofit METR work inside its offices and publish what they find without Anthropic's approval.