Anthropic proposes US AI bill limiting intervention to 50 injuries or €920 million in damage
Anthropic has circulated a draft bill to US lawmakers that would let the government block an advanced AI model only if it poses an imminent risk of killing or seriously injuring more than 50 people, or causing over €920 million (US$1 billion) in damage. According to POLITICO, the risk would have to involve a biological weapon, a model escaping its developer’s controls or a cyberattack; the draft would also require independent evaluations and give the Commerce Department sole authority to seek a halt from a judge.
Bottom line — The draft would set a federal intervention threshold while overriding some state AI rules; Congress has not passed it.
Go deeper 6
-
Anthropic told POLITICO the proposal reflects its Advanced AI Framework, which it has promoted to federal and state lawmakers since June.
-
Under the draft, evaluators finding an imminent risk covered by the bill would have to notify the Commerce Department within 72 hours, according to PYMNTS.com.
-
NewsBytes reports that AI companies could pay for independent evaluations, but payment could not depend on the outcome.
-
According to India Today, companies would also have to publish risk reports, set out safety plans, report serious incidents and protect employees who raise safety concerns.
-
The draft resembles California’s SB 53, POLITICO reported, but would override some state requirements.
-
Charlie Bullock of the Institute for Law & AI called the proposal reasonable but weaker than the competing FRONTIER Act, WION reported.