U.S. Treasury Secretary Scott Bessent has urged AI companies to assume responsibility for their own systems rather than depending on government regulation. In an interview with CNBC, he stated, “it is humans who are responsible, not the AI,” and dismissed the idea of granting companies legal protections against liability. His remarks came as calls to establish guardrails related to the development of the technology mount.
The Trump administration has shown resistance to proposals aimed at slowing AI advancements, with President Donald Trump insisting the U.S. must preserve its technological lead, especially in relation to China. He recently proposed establishing an “AI Force,” structured similarly to his earlier Space Force initiative, though the new unit would focus specifically on AI leadership. The administration has yet to provide specifics about the task force’s objectives or composition.
Executives in the AI industry have begun advocating for greater caution in development. Dario Amodei, CEO of Anthropic, published an essay entitled “We Must Pace The Frontier” arguing that rapid expansion of advanced AI systems must slow to allow safety protocols to catch up. He warned that if AI innovation is not intentionally moderated, AI agents could become capable of “taking over the entire internet” within 6-12 months. Both Elon Musk, CEO of Tesla and SpaceX, and Sam Altman, CEO of OpenAI, publicly endorsed Amodei’s calls for moderation.
Read Also: Vital Skydance Nears Warner Bros. Acquisition, Shares Surge
A coalition of over 100 AI safety researchers has also pushed for stricter regulatory measures, insisting that external evaluations must operate independently to be effective. The AI Evaluator Forum released a statement outlining necessary conditions for credible third-party assessments, including unrestricted system access, protection from company retaliation, and full editorial control over findings. The group emphasized that oversight mechanisms cannot function properly if evaluators lack autonomy or face conflicts of interest.
The proposed oversight framework would mandate that companies provide evaluators with full access to their systems, datasets, and personnel, as well as direct lines of communication with executive leadership. Without these guarantees, the experts warned, external reviews would lack credibility or become easily manipulated by corporate interests.