Donald Trump said last week he viewed AI startup Anthropic as a national security threat, but his tone has softened markedly after direct engagement with the company's leadership, he revealed in an exclusive interview with Axios.
The president told Axios' Marc Caputo at the White House that his concerns have eased. "Well, not now, but a week ago, maybe," Trump said when asked if he saw Anthropic or CEO Dario Amodei as a national security risk.
Trump credited Amodei's quick response to government demands for the thaw. "He responded to us very quickly because you know it's a tremendous liability," the president said. "People get put in prison immediately for that. You can't play games with that. And he responded very responsibly, I thought." The remarks suggest the company understood the gravity of the situation it faced.
The tensions erupted after Amazon flagged a vulnerability in Anthropic's most advanced models. The Commerce Department subsequently restricted access to those systems for foreign nationals and countries outside the United States. The Pentagon had also labeled the company a supply chain risk, subjecting it to treatment normally reserved for foreign adversaries.
Behind the scenes, the administration felt that Anthropic initially dismissed its concerns during early conversations. Technical discussions have since unfolded in Washington, with both sides now working on standards to evaluate AI security breaches. The episode underscored how Anthropic has had to recalibrate its communications with government officials as much as address technical questions.
Trump did not entirely rule out wielding the Defense Production Act to force Anthropic's compliance, though he signaled restraint. "I have the power to use a lot of things," he said. "But I'm not sure I have to do that."
The president indicated that Amazon's role in raising alarms was partly motivated by competitive tension. "It was a competitor and a part owner that turned Anthropic in," Trump said. "They didn't like what they were doing. They were very concerned." He added that Anthropic has since "been very responsible."
Strategic competition with China, not punishment of domestic AI firms, remains Trump's broader priority. He said the United States is winning the AI race decisively and that shutting down Anthropic would be self-defeating. "I was with President Xi. We talked about it. We're beating China by a lot," Trump said. "The good far outweighs the bad. We are going to find the bad and we're going to stop it."
Anthropic responded with a statement expressing gratitude for the administration's partnership. "We remain committed to working alongside them towards our shared goals of protecting critical infrastructure and making sure the U.S. leads in AI," the company said.
The thaw in relations masks deeper uncertainties. While Trump and Amodei appear to have found common ground on immediate concerns, the technical work of setting AI safety standards and determining the scope of international cooperation remains unresolved.
Author James Rodriguez: "Trump's sudden shift from threat assessment to bromance suggests the real issue was never ideology but communication, and perhaps Amazon's competitive knife in Anthropic's back taught the startup a brutal lesson in Washington's game."
Comments