'Not now, but a week ago, maybe': Trump says Anthropic not a national security threat

US President Donald Trump, in an interview with Axios, said that he did not view Anthropic or its CEO Dario Amodei as a threat to national security. In the Axios Show, Trump said, “Well, not now, but a week ago, maybe.” Trump claimed that following the lunch in the resort city of Évian-les-Bains, between the G7 leaders and leading Silicon Valley and global AI executives, he left with the impression that Amodei was "nice" and "smart."

"He responded to us very quickly because you know it's a tremendous liability," Trump said to the Axios Show. "People get put in prison immediately for that. You can't play games with that. And he responded very responsibly, I thought."

Trump responded to the question of whether he would leverage emergency powers under the Defence Production Act if the defence labs did not get in line. "I have the power to use a lot of things," said Trump, "But I'm not sure I have to do that."

Also Read: Who gets to define terrorism? How colonial history shapes the debate over Hezbollah's designation

Last week, the US imposed an export control on its most powerful AI models, Fable 5 and Mythos, on foreign nationals both inside and outside the US over the potential of “jailbreak”.

"We are grateful to the administration for their ongoing partnership in working to get this matter resolved as quickly as possible," said Anthropic in a statement.

It was discovered by Amazon security researchers and reported to the Trump administration. Anthropic responded by blocking access to Claude Mythos and Fable 5. It claimed no tester has been able to provide a “universal jailbreak”, something that can broadly bypass all the safeguards and unblock a wide range of cyber capabilities. It also asserted that perfect jailbreak resistance is not available. Some minor vulnerabilities are likely to remain.

"It was a competitor and a part owner that turned Anthropic in. They didn't like what they were doing. They were very concerned," Trump said of the concerns raised by Amazon. “I think so far it's been very responsible.”

The Pentagon has already designated Anthropic as a supply chain risk. Despite a temporary restraining order issued by a federal judge in late March, the Pentagon's 180-day timeline to completely remove Anthropic's Claude models from all Department of War systems is still moving forward.