Trump and Bessent Reject Federal AI Regulations for Self-Policing

President Donald Trump and Treasury Secretary Scott Bessent have rejected requests from AI industry leaders for new federal regulations to mitigate existential risks, characterizing such warnings as a hoax and alarmist. During a meeting with executives from OpenAI, Microsoft, Meta, and SpaceXAI, Trump advocated for industry self-regulation and the use of outside auditors to assess safety rather than government mandates.

Treasury Secretary Scott Bessent further criticized tech leaders who seek government guardrails, comparing their requests to the fictional character Hannibal Lecter by suggesting they are asking the state to stop them before they cause harm. Bessent argued that labs must accept responsibility for their own risks to ensure safe acceleration and maintain the United States' technological lead over China. While Bessent praised the cooperation of labs and cited OpenAI's decision to shelve a new model as a positive step, the approach faces internal industry opposition. Anthropic CEO Dario Amodei has called for government regulation, and OpenAI safety leader David Robinson resigned on October 3, claiming the company is moving too fast to ensure societal safety.

To manage global stability, the United States and China have agreed to establish a US-China AI dialogue and a bilateral notification system to share warnings about serious incidents, such as biological threats or uncontrolled AI agents. This agreement follows discussions between Bessent, Chinese Vice Premier He Lifeng, and President Xi Jinping.