Thune claims Trump is not dug in against AI guardrails

Senate Majority Leader John Thune says President Donald Trump is "not really dug in" against implementing guardrails for artificial intelligence, a position that contrasts with the president's public opposition to restricting the technology. Thune told Axios that he believes some of Trump's strong pushback is directed at what he described as public "paranoia and noise around the subject" rather than the underlying policy merits.

Thune, who spoke to the outlet on Wednesday at his Capitol office, said he had a private conversation with Trump a few weeks ago as the AI issue was becoming "pretty hot." The Senate leader interprets the president's stance as a preference for technology companies to develop their own methods for securing systems, rather than having Congress or the administration impose strict regulations. This approach would rely more heavily on experts to provide guidance. Trump has previously characterized recent fears surrounding AI safety as a "hoax," but Thune stated, "I do think that the threats are real."

The legislative landscape remains complicated by conflicting signals from both the White House and the tech industry. Thune noted that he has Trump's skepticism in one ear and alarm from tech leaders in the other. Dario Amodei, the CEO of Anthropic, called Thune to discuss AI safety before issuing his latest public warnings about the risks of the technology. The two executives also discussed Thune's proposed AI bill during that conversation.

Thune referenced legislation he has introduced in past years with Senator Amy Klobuchar (D-Minn.), which is not yet reintroduced in the current Congress. The bill places the responsibility on tech companies to mitigate AI risks and requires them to report on serious risks associated with frontier models. It also mandates that companies develop plans to address these concerns. Thune stated, "I think you have to have at least some sort of a plan and some guardrails around it, which is what I think our bill does." However, he added that he believes there are individuals engaging in "a lot of fear mongering."

Klobuchar is currently seeking additional Democratic co-sponsors to re-introduce the measure, including Senators Chris Coons (D-Del.) and Maria Cantwell (D-Wash.). According to Thune, Democrats have two primary objections to the current bill language. First, pre-deployment testing is optional for companies under the proposed rules, whereas Democrats want it mandated. Second, Senator Cantwell is requesting that the Department of Commerce be granted AI rulemaking authority. Thune expressed concern that granting such authority could give the executive branch too much power.

Thune possesses specific experience in this policy area, noting that if he had not become Senate Majority Leader, he would have again chaired the Commerce committee, which holds jurisdiction over much of the AI regulatory framework.