Anthropic quietly pitches its own AI safety bill to lawmakers, Politico reports

People visit an Anthropic booth at Moscone Center during the Dreamforce 2026 technology summit in San Francisco, California, US, September 17, 2026. REUTERS/Carlos Barria/File Photo

Oct 16 (Reuters) - Anthropic representatives ⁠have quietly pitched Capitol Hill lawmakers on a draft bill ⁠to regulate frontier AI safety and security, Politico reported on ‌Friday, citing three people with knowledge of the efforts.

The bill would let the government block a model's release or use only if it poses a foreseeable, imminent ​risk of more than $1 billion in damages, ⁠or of killing or seriously ⁠injuring more than 50 people, according to the report.

Here are some ⁠details:

• ‌The risk must stem from the creation or release of a biological weapon, the model evading its developer’s controls ⁠or a cyberattack, Politico reported

• Anthropic and the Department ​of Commerce did ‌not immediately respond to Reuters requests for comment

• Under the ⁠bill, independent ​evaluators, which would register with the Commerce Department, would be tasked with periodically evaluating the lab and its frontier models, according to the report

• ⁠The labs could pay these evaluators, provided ​the payment is not tied to the evaluation's outcome, the report said

• If the evaluators determine there is “imminent enumerated risk” for a company’s frontier ⁠models, they must report it to Commerce within 72 hours of completing their assessment, Politico reported

• While the Commerce Department could request a judge to halt a model after determining risk, the language ​explicitly prohibits any other US government department, agency, ⁠or officer from suspending development, deployment, or internal use, according to ​the report

• A federal judge for the ‌District Court for the District of Columbia ​could order the suspension of a model, the report said

(Reporting by Jaspreet Singh in Bengaluru; Editing by Anil D'Silva)

Thank you for your report!

2026年9月17日，在美国加利福尼亚州旧金山举行的Dreamforce 2026科技峰会期间，人们在莫斯康展览中心参观Anthropico展位。（路透社/卡洛斯·巴里亚/资料照片）

10 月 16 日（路透社）——据 Politico 周五报道，Anthropico 的代表已悄悄地向国会山议员们推介了一项旨在监管前沿人工智能安全保障的法案草案，该报道援引了三位知情人士的话。

据报道，该法案将允许政府在模型发布或使用存在可预见的、迫在眉睫的风险，造成超过 10 亿美元的损失，或导致 50 人以上死亡或重伤的情况下，阻止该模型的发布或使用。

以下是一些细节：

据Politico报道，风险必然源于生物武器的制造或释放、该模型逃避了开发者的控制，或是网络攻击。

• 安特罗皮克公司和美国商务部均未立即回应路透社的置评请求。

• 据报道，根据该法案，独立评估机构（需在商务部注册）将负责定期评估该实验室及其前沿模型。

报告称，实验室可以向这些评估人员支付报酬，前提是报酬与评估结果无关。

据Politico报道，如果评估人员认定某公司的前沿模型存在“迫在眉睫的列举风险”，则必须在完成评估后的72小时内向商务部报告。

• 报告指出，虽然商务部可以在评估风险后请求法官叫停某个模型，但相关条款明确禁止任何其他美国政府部门、机构或官员暂停该模型的开发、部署或内部使用。

报道称，哥伦比亚特区地方法院的一名联邦法官可以下令暂停模特的职业资格。

（贾斯普雷特·辛格在班加罗尔报道；阿尼尔·德席尔瓦编辑）

感谢您的报告！