Here is something that would have been unthinkable six weeks ago: President Trump this morning told CNBC that Anthropic is "shaping up" and that a deal for the company's AI to work inside the Department of Defense is "possible."
That is a significant reversal.
In March, the Pentagon declared Anthropic a supply chain risk and required defense contractors to certify they had stopped using Claude in any work with the military. Trump then went further, posting on Truth Social that his administration would "IMMEDIATELY CEASE all use of Anthropic's technology" and would "not do business with them again."
Anthropic sued. A federal judge temporarily blocked the order. And Claude, widely used by defense contractors and federal agencies, kept running anyway.
Then last Friday, Anthropic CEO Dario Amodei walked into the White House to brief senior officials on Mythos, the company's powerful new private AI model. White House Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent were in the room. The meeting was described as "productive and constructive."
This morning, Trump called Anthropic "very smart" and said they can "be of great use." His tone, on television, was a pivot.
Why This Matters to Small Business Owners
This is not just a Washington story. It is a business reliability story.
Claude is one of the three most widely used AI platforms by small businesses, alongside ChatGPT and Gemini. Hundreds of thousands of entrepreneurs use it daily: for drafting proposals, handling customer service, building internal tools, processing documents, and running AI agents that automate operations.
When the Pentagon blacklisted Anthropic in March, it created something small business owners rarely think about: AI platform risk. The question was not just whether the federal government would use Claude. It was whether the corporate environment would follow. Large clients, government contractors, and compliance-sensitive industries began asking whether they should keep using a platform that had been labeled a national security concern.
That uncertainty is now starting to lift.
If a deal gets done between Anthropic and the Pentagon, the blacklist disappears. The "supply chain risk" label goes away. The lawsuit settles. And Anthropic goes from embattled AI company to preferred defense technology partner, practically overnight.
For small businesses that rely on Claude, that is a very different risk environment than existed 30 days ago.
What Changed
Two things happened that shifted the dynamic.
First, Anthropic unveiled Mythos, a model that by multiple accounts is the most capable AI coding and reasoning system currently available. Sergey Brin told DeepMind employees this month that Google needs to catch up to Anthropic. The NSA reportedly already has access to it. The model is powerful enough that the government's desire to use it appears to have outweighed its desire to punish the company.
Second, the legal position changed. A federal judge blocked Trump's order. That gave Anthropic leverage and gave both sides a reason to talk rather than fight.
The result: Dario Amodei in the White House, Trump calling them "very smart" on Squawk Box, and a likely path toward resolution.
What to Watch For
A formal resolution to the DoD blacklist would come through either a court settlement or an executive reversal. Neither has happened yet. Trump's comments this morning are a signal, not a done deal.
But the direction is clear.
For small business owners who paused Claude adoption over the past six weeks because of the controversy, or who faced pushback from clients or compliance teams, the next few weeks should bring clarity. A deal would restore confidence in Anthropic as a stable, government-acceptable AI vendor.
Until then, Claude keeps running. The federal court order keeps it available. And this morning, the President of the United States said the company is shaping up.
That is about as good a signal as Anthropic could have asked for.
Sources: CNBC, April 21, 2026 | The Verge, April 21, 2026