WASHINGTON — Treasury Secretary Scott Bessent defended the Trump administration’s hands-off approach to artificial intelligence Tuesday, telling lawmakers that companies developing the most advanced systems can stop their work without federal intervention. “They could stop any time they want to,” Bessent said during testimony before the House Financial Services Committee. “They could stop their progress any time they want to.” Rep. Maxine Waters of California, the committee’s top Democrat, questioned Bessent about warnings from technology executives who have urged a pause in the development of so-called frontier AI models. Waters accused President Donald Trump of promoting AI and data centers without adequate protections as threats to the public increase. “President Trump continues to blindly push for AI and data centers with absolutely no safeguards,” Waters said. The hearing followed a weekend push by executives at major U.S. AI companies for a slowdown, citing concerns about the possibility that increasingly capable systems could pose existential risks. Their calls have intensified a debate over whether private companies should be allowed to set their own safety limits or whether the federal government should impose enforceable rules. Bessent said the administration has already begun reviewing advanced AI models, saying that work started more than five or six months ago. He also has been involved in efforts to protect financial institutions from AI-enabled cyberattacks and other threats that could bypass existing defenses. The Treasury Department is operating an information-sharing clearinghouse for financial companies, allowing them to exchange strategies for strengthening their cybersecurity against advanced AI systems. Trump has rejected recent warnings about AI as a “hoax” and opposed growing demands for additional guardrails. The president and Bessent have instead presented AI regulation as part of a strategic contest with China, arguing that allowing Beijing to gain an advantage could endanger the United States. “There is no day after tomorrow if China wins at this,” Bessent said last week. Artificial intelligence is expected to be discussed when Trump meets Chinese President Xi Jinping later this month. Before that summit, Bessent plans to meet this weekend with Chinese Vice Premier He Lifeng. Bessent has said he oversees the administration’s diplomatic and economic engagement with China on AI. The administration’s position leaves major decisions about the pace and safety of AI development largely with the companies building the technology, even as those companies’ own leaders warn that voluntary restraint may be necessary to address risks affecting the public, the economy and national security.