Democratic Senators Demand Disclosure on Tech Giants' Influence Over AI Regulation
In a letter made public on Tuesday, October 6, Democratic Senators Elizabeth Warren and Richard Blumenthal pressed Trump administration officials, including Treasury Secretary Scott Bessent and Commerce Secretary Howard Lutnick, to disclose whether technology giants have used lobbying to shape the government's AI regulatory policy and to reveal details of the regulatory framework and testing mechanisms.
The senators accused the administration of "catering to the interests of big tech CEOs" rather than addressing the safety and national security risks posed by increasingly powerful AI systems. The letter focuses on the administration's revision of the AI regulatory framework between May and June. According to the senators, the Office of the National Cyber Director (ONCD) initially drafted a plan requiring mandatory 60-day government evaluations of frontier AI models, followed by a 75-day restricted user testing period, with the ability to block deployment of models deemed to pose unacceptable risks. After opposition from executives at Google, Anthropic and OpenAI, the mandatory framework was reportedly changed to a voluntary mechanism with a narrower scope. Following concerns raised by Meta CEO Mark Zuckerberg and xAI CEO Elon Musk, the testing window was further cut from 90 days to 30 days, and the final version was incorporated into Executive Order 14409 signed in June.
The senators also criticized the White House Accord on Super Intelligence as an industry-led self-regulation arrangement, citing reports that Zuckerberg played a "central role" and that Nvidia CEO Jensen Huang helped rally support. They argued the accord cannot replace binding government regulation.
Warren and Blumenthal demanded a response by October 19, including records of meetings and communications between tech companies and government officials regarding Executive Order 14409, different draft versions, and a 31-question survey sent by ONCD to industry players along with company responses. They also asked whether independent evaluators, consumer organizations, critical infrastructure operators and academic experts were consulted, and what capabilities and risks the current pre-deployment tests assess, as well as what steps the government would take if models fail.
The letter is the latest in a series of pressure from the two senators on the administration's AI policies. On September 28, they asked the Treasury Department to explain the secret, voluntary testing mechanism for advanced AI models. Meanwhile, Republican Senator Bernie Moreno sent a letter to Anthropic CEO Dario Amodei criticizing an "alarmist" approach to AI risks, underscoring the partisan divide in Washington over AI policy.
