Trump Opens Door to Pentagon Deal With Anthropic After White House Talks
President Donald Trump said Tuesday that Anthropic is "shaping up" and signaled openness to reversing the Pentagon's blacklisting of the AI company, marking a potential shift after months of tension between the administration and the San Francisco lab.
Trump directed the government to stop working with Anthropic in February. The Pentagon followed by declaring the firm a supply-chain risk, blocking Defense Department employees and contractors from using its tools after a six-month transition period.
The move came after Anthropic sought Pentagon assurances that its Claude AI models would not be used for surveillance of Americans or autonomous weapons operations. Anthropic filed suit against the Defense Department in March, disputing the blacklist determination.
Anthropic CEO Dario Amodei met with White House officials Friday. Trump told CNBC the meeting went well. "They came to the White House a few days ago, and we had some very good talks with them," he said. "I think they're shaping up. They're very smart, and I think they can be of great use."
Asked about a Pentagon deal, Trump said: "It's possible. We want the smartest people."
Anthropic said its White House meeting focused on cybersecurity, American competitiveness in AI, and AI safety. The company did not comment further on Trump's remarks.
Mythos and Project Glasswing
The timing matters. Anthropic unveiled Mythos, its most advanced model, weeks before the White House meeting. Security experts said the tool has an unusual ability to identify cybersecurity vulnerabilities and devise ways to exploit them.
Anthropic will not release Mythos publicly. Instead, it launched Project Glasswing, inviting major tech companies, cybersecurity vendors, JPMorgan Chase, and dozens of other organizations to privately test the model and prepare defenses.
Anthropic co-founder Jack Clark said last week the company was discussing Mythos with the Trump administration but provided no details.
Legal and Political Status
A federal appeals court in Washington declined this month to block the Pentagon's blacklist, backing the Trump administration. Another appeals court had reached the opposite conclusion in a separate legal challenge by Anthropic.
Trump still called Anthropic's staff "the radical left" in his Tuesday comments, even while signaling a thaw. The company's push for guardrails on military AI use remains a point of friction with an administration that has prioritized speed in defense applications.
For government employees and contractors, the blacklist's potential reversal would restore access to AI for Government tools that agencies have found valuable, particularly Claude's coding capabilities. Any deal would likely require Anthropic to agree to Pentagon terms on how its models are deployed.
Your membership also unlocks:
