President Donald Trump revealed that he recently viewed the AI company Anthropic as a national security threat, according to an exclusive interview for “The Axios Show.” However, the president also indicated that relations between the White House and the company have improved in the days since.

Why It Matters

The episode placed Anthropic — one of the heavyweights in artificial intelligence — squarely in the middle of a government crackdown carrying both domestic and international consequences. A combination of national security concerns and personality clashes drove the conflict.

The treatment Anthropic faced was striking in its severity. Between the Commerce Department imposing sweeping export controls and the Pentagon designating the company a supply chain risk, Anthropic was subjected to measures typically reserved for foreign adversaries.

What Trump Said

During a wide-ranging White House interview, Axios’ Marc Caputo asked Trump directly whether he considered Anthropic, or its CEO Dario Amodei, a threat to national security.

The president’s response was telling. He said that while he didn’t view the company that way now, he might have a week earlier. Yet Trump also offered a softer assessment of Amodei following the G7 summit, describing him as “nice” and “smart.”

Trump credited the CEO with responding swiftly to the administration’s concerns, framing the matter as a serious liability. He suggested that such situations carry grave consequences and praised Amodei for handling the issue responsibly.

How the Conflict Unfolded

The friction escalated quickly. Last week, the Trump administration restricted access to Anthropic’s most advanced models for any country outside the United States, as well as for foreign nationals within the country.

The trigger appears to have been a report from Amazon detailing a vulnerability, which alarmed the administration. When officials raised the issue with Anthropic’s leadership, they reportedly felt dismissed — fueling the standoff.

What followed was a series of technical discussions in Washington. According to the report, the two sides are now working together on standards to evaluate AI “jailbreaks,” or attempts to bypass an AI system’s safeguards. For Anthropic, much of the challenge has reportedly been learning how to communicate effectively with the administration, as much as reaching a shared understanding of how the technology works.

The Threat of Emergency Powers

Despite signs of a thaw, Trump did not rule out using emergency authority under the Defense Production Act (DPA) if the company failed to fall in line — a tactic that had previously been threatened during the dispute with the Pentagon.

Asked about the possibility, Trump acknowledged having broad powers at his disposal but said he wasn’t sure he would need to use them. He also revealed an unusual detail about how the concerns first surfaced, claiming that a competitor and part-owner had effectively reported Anthropic, citing serious worries about the company’s activities. Even so, Trump said the company had so far behaved responsibly.

Anthropic’s Response

For its part, Anthropic struck a conciliatory tone. In a statement, the company expressed gratitude to the administration for its ongoing partnership in working to resolve the matter as quickly as possible.

The company emphasized its commitment to collaborating with the government toward shared goals — namely, protecting critical infrastructure and ensuring that the United States maintains its lead in artificial intelligence.

The Bigger Picture: Beating China

Underlying the entire dispute is a broader strategic priority. In Trump’s view, the race to outpace China on AI ultimately outweighs his political clashes with Anthropic and its competitors.

Trump made clear he has no desire to shut down Anthropic, pointing to America’s strong position in the AI race. He recounted discussing the issue directly with Chinese President Xi Jinping, asserting that the United States is beating China “by a lot.” In his framing, the benefits of American AI leadership far outweigh the risks — though he vowed to identify and stop any problems that arise.

The Bottom Line

The relationship between the White House and Anthropic appears to be on the mend after a turbulent stretch. Still, significant uncertainty remains. The harder work — establishing meaningful AI safety standards and defining what international cooperation should look like — is far from settled, leaving the long-term path forward unclear.