According to TechCrunch, Anthropic has been shopping around a draft AI safety bill that it hopes Congress will consider adopting, as recently reported by Politico.
The proposed legislation would allow blocking the release of AI models that pose a “foreseeable, imminent risk” of causing over 50 deaths or more than one billion dollars in property damage, according to Politico. The catastrophic risk must “stem from the creation or release of a biological weapon, the model evading its developer’s controls or a cyberattack,” TechCrunch reports.
These thresholds mirror those in California’s recently passed Senate Bill 53, which similarly defined such scenarios as “catastrophic risk,” according to TechCrunch.
Under Anthropic’s proposal, “independent evaluators” registered with the U.S. Commerce Department would assess models for potential disasters, TechCrunch reports. According to Politico, AI labs could potentially pay these evaluators “so long as the payment is not contingent on the evaluations’ results,” though it remains unclear how this system would ensure payment independence.
When evaluators identify an imminent catastrophic risk, they would have 72 hours to notify the Commerce Department, which would then need to request a Washington, D.C., judge to halt the model’s release, according to TechCrunch. The legislation would prohibit federal agencies from halting distribution on their own discretion, Politico reports.
An Anthropic spokesperson confirmed the draft bill to Politico, stating it represents proposals from Anthropic’s Advanced AI Framework.