> ## Content Index
> Fetch the complete content index at: https://wire.fourthweb.ai/llms.txt
> Use this file to discover other available public pages before exploring further.

# Trump Declares He "Likes" the AI CEO He Once Called a Security Risk
- URL: https://wire.fourthweb.ai/trump-declares-he-likes-the-ai-ceo-he-once-called-a-security-risk/
- Published: 2026-10-01T12:18:57.000Z
- Updated: 2026-10-01T13:00:45.000Z
- Description: The president who threatened to boot AI companies from federal contracts just had dinner with the CEO he called a "security risk" — and now he "likes" him.
- Author: Travis Wright
- Tags: AI Agent Economy, AI Agents, OpenAI, Anthropic, Google AI, Circle

**The president who threatened to boot AI companies from federal contracts just had dinner with the CEO he called a "security risk" — and now he "likes" him.**

### The Summary

- [Trump said he "liked" Anthropic CEO Dario Amodei after a recent dinner](https://www.bloomberg.com/news/articles/2026-10-01/trump-says-he-liked-anthropic-s-amodei-in-sign-of-detente?ref=wire.fourthweb.ai), a sharp reversal from the Defense Department's supply chain risk label on the company
- This signals potential thawing of tensions between the White House and one of the three frontier AI labs, after months of regulatory pressure on Claude's maker
- The timing matters: [Anthropic](https://wire.fourthweb.ai/tag/anthropic/) is mid-negotiation on federal AI contracts worth hundreds of millions, and Trump's social signal could unlock doors previously slammed shut

### The Signal

The Defense Department labeled Anthropic a supply chain risk in July 2026\. Three months later, Trump is posting about liking Dario Amodei. That's not policy whiplash, that's strategy.

[The dinner apparently happened last week](https://www.bloomberg.com/news/articles/2026-10-01/trump-says-he-liked-anthropic-s-amodei-in-sign-of-detente?ref=wire.fourthweb.ai), though neither side has disclosed the guest list or location. What we know: Trump went from threatening to exclude AI companies with foreign investment ties from federal work to breaking bread with the CEO of a company backed by $7.3 billion from investors including Google and Spark Capital. Anthropic's cap table includes non-U.S. entities. That was supposedly the problem.

> "Trump went from threatening exclusion to breaking bread in 90 days."

But here's what changed. [OpenAI](https://wire.fourthweb.ai/tag/openai/) locked in a $15 billion Department of Defense contract in August. [Google DeepMind](https://wire.fourthweb.ai/tag/google-ai/) signed a $9 billion multi-year federal AI infrastructure deal in September. Anthropic was on the outside, watching competitors absorb government budgets while DOD lawyers circled their corporate structure with red pens. Now Trump says he "likes" Amodei. Translation: the door just cracked open.

**Key dynamics at play:**

- Anthropic trails OpenAI and Google in federal AI revenue by roughly $22 billion over the next three years
- The company has repositioned Claude as the "constitutional AI" model, emphasizing safety and interpretability, exactly the language federal procurement offices use
- Amodei has been quieter than Altman or Hassabis on AGI timelines and existential risk, making him an easier dinner guest for a president allergic to alarmism

The timing also matters for Anthropic's enterprise strategy. They launched Claude for Government in June, specifically targeting federal agencies with air-gapped deployments and FedRAMP compliance. But you can't sell to the government if the Defense Department has you on a watch list. One dinner, one Trump post, and suddenly that FedRAMP certification has teeth again.

### The Implication

If you're watching the agent economy, watch procurement. Federal AI budgets will top $50 billion annually by 2028, and the companies that win those contracts will set the standards everyone else follows. Anthropic just went from blacklist to maybe-list. That's not just good for their revenue, it's a signal about what kind of AI the government wants to buy.

For everyone else building agents: constitutional AI and interpretability just became competitive advantages, not nice-to-haves. If Trump is willing to flip on Anthropic because they framed their models the right way, other buyers will follow. The age of "move fast and break things" is over. The age of "move fast and explain things" just started.

### Sources

[Bloomberg Tech](https://www.bloomberg.com/news/articles/2026-10-01/trump-says-he-liked-anthropic-s-amodei-in-sign-of-detente?ref=wire.fourthweb.ai)