Anthropic CEO Dario Amodei walked into the West Wing on Friday for a sit-down with White House chief of staff Susie Wiles. Let that sink in for a second.

This is the same administration that blacklisted Anthropic as a "supply chain risk." The same Pentagon that told every federal agency to stop using Claude. The same White House whose appeal against a judge's injunction protecting Anthropic is still winding through the courts.

And now the chief of staff is taking the meeting. First reported by Axios and confirmed by CNN, Reuters, and the Washington Examiner, the Friday sit-down represents what sources close to the negotiations describe as a "breakthrough" in the company's bitter dispute with the Trump administration over military AI.

What changed

One word: Mythos.

When Anthropic unveiled Claude Mythos on April 7th, the administration suddenly had a problem. The model it had blacklisted turned out to be the most advanced cybersecurity AI ever built. Treasury Secretary Bessent called it a "step function change in abilities" at a Wall Street Journal event. That triggered an emergency meeting with Wall Street CEOs and a separate briefing that included Vice President Vance and Elon Musk.

The model can scan software, detect serious security flaws, and figure out how those flaws could be exploited, often with minimal human input. Five central banks have issued warnings. Fourteen finance ministers held an emergency meeting. The intelligence community, CISA, and Treasury are all already testing it.

Put simply: the government banned the company, then realized it could not defend itself without the company's product.

The red lines

The original dispute came down to two things Anthropic refused to do. Its models would not be used for mass domestic surveillance. And they would not power fully autonomous weapons systems. These have been core to the company's identity since Amodei and his team left OpenAI in 2021 specifically because they believed AI development needed stronger ethical boundaries.

The Pentagon wanted unrestricted access. Anthropic said no. The Pentagon blacklisted them. Anthropic sued. A federal judge granted a preliminary injunction in March. The Trump administration appealed.

Meanwhile, Anthropic pivoted to a parallel political strategy, hiring Ballard Partners, a lobbying firm with deep ties to Trump's inner circle. That's how you get from "blacklisted" to "West Wing meeting" in six weeks.

The leverage play

A source close to negotiations gave Axios a line that tells you everything about the power dynamic: "It would be grossly irresponsible for the U.S. government to deprive itself of the technological leaps that the new model presents. It would be a gift to China."

That is not the language of a company begging for mercy. That is the language of a company that knows it has something Washington cannot build or buy from anyone else. And it knows Washington knows.

What this means

If you have been following this saga on The AI Post, you already know the pattern. Blacklist. Secret testing. OMB memos opening the door. And now: the meeting that makes it official.

This is Amodei's second high-stakes encounter with a senior Trump official this year. If Friday goes well, expect the "supply chain risk" designation to quietly disappear, the appeal to be dropped, and Anthropic to emerge as the most politically connected AI company in America. The safety company that refused to build weapons may have just won the biggest government contract in AI history by saying no.

Sometimes the best negotiating position is having something they cannot get anywhere else.

First reported by Axios. Confirmed by Reuters, CNN, and the Washington Examiner.