U.S. President Donald Trump said he no longer views Anthropic as a national security threat. On June 19, local time, blockchain outlet Cryptopolitan reported that Trump made the remark after meeting Anthropic Chief Executive Dario Amodei (다리오 아모데이) at a G7 summit.

Asked whether Anthropic or Amodei was a threat, Trump replied that it was not now, but it might have been a week earlier. He also said he could invoke the Defense Production Act but did not see the need to do so.

Existing regulations remain in place. Under a measure the Commerce Department took on June 12, foreigners must receive federal approval to access Mythos 5 and Fable 5. A supply-chain risk designation issued by the Defense Department on March 3 has also not been withdrawn, and a ban on federal agencies using Anthropic technology continues.

The shift in tone was prompted by a G7 luncheon held in Evian-les-Bains, France, on June 17. Amodei, together with Demis Hassabis, proposed a U.S.-led AI alliance. The plan calls for democratic countries to jointly coordinate advanced AI trade and standards while excluding China. Amodei also urged countries not to split their approaches to AI regulation.

After the luncheon, Trump described Amodei as smart and likeable. The meeting was their first face-to-face encounter since the March supply-chain risk designation. Earlier, Trump’s February directive on Truth Social had described Anthropic executives as ideologically hostile figures to the Defense Department.

Amazon’s vulnerability report was cited as the direct trigger for the Commerce Department measure. Trump explained that behind Commerce Secretary Howard Lutnick contacting Amodei was an issue raised by a company that is both an Anthropic investor and a competitor. Amazon has invested $8 billion in Anthropic and also runs competing models on AWS Bedrock. After being notified on June 12, Anthropic had to respond within 90 minutes and judged it could not reliably apply access controls by nationality, so it halted the two models for all users, including employees outside the United States.

The regulations are also weighing on listing plans. Anthropic completed its Series H-1 fundraising at the end of May with an investment amounting to $65 billion and was assigned a corporate value of $965 billion. It then confidentially submitted an S-1 to the U.S. Securities and Exchange Commission on June 1, and annual recurring revenue was tallied at about $47 billion in early June. With its core Mythos and Fable lines now subject to controls, the longer the regulations last, the greater the burden could become on valuation for the listing.

The matter is also affecting policy responses by allies. Canadian Prime Minister Mark Carney said the restrictions show the need to build and diversify AI supply chains. Japan also said in a policy draft released on June 19 that it would actively and continuously review its AI legal framework.