The most cautious voice in AI just had dinner with the man who thinks safety talk is for losers, and the collision tells you everything about who's actually steering the ship.
The Summary
- President Trump rejected new federal AI regulations after meeting with industry titans including Nvidia's Jensen Huang, Anthropic's Dario Amodei, and Palantir's Alex Karp at the White House.
- The meeting came as fresh AI model breaches intensified global safety concerns, with "a far wider range" of advanced model failures than previously disclosed.
- A memo attacking Amodei circulated inside the White House before his one-on-one dinner with Trump, after Trump mocked him on Truth Social for calling AI companies to slow down.
- America First meets AI safety: Trump's "let's go and let's win" stance against China now determines whether frontier models get guardrails or just get shipped.
The Signal
Anthropic CEO Dario Amodei walked into the White House for a Sunday dinner with President Trump representing the careful wing of AI development. He's publicly called for "pacing" the rollout of the most advanced models and stronger safety guardrails. Trump has spent weeks dismissing that position, calling Amodei and others AI "people" who want to do "bad, or potentially bad, things" by slowing American progress. After that dinner and a follow-up lunch with Huang and Karp, Trump shot down the idea of new federal AI regulations entirely.
The timing matters. New disclosures about AI model breaches dropped just before these meetings, revealing that advanced AI systems are failing in ways nobody previously reported publicly. Global anxiety about AI risk is climbing. Amodei showed up with the receipts for why guardrails matter. Trump showed up with a different agenda: beat China, move fast, don't let regulation slow American companies.
"President Trump has been clear: America will lead the world in Super Intelligence, while protecting American consumers."
The White House framed the meetings around competitiveness and innovation, not safety reviews or regulatory frameworks. Fortune reported that someone inside the administration circulated a memo attacking Amodei before the dinner. That's not the behavior of an administration looking for honest counsel on risk. That's an administration looking to neutralize dissent before it can take root.
The cast list tells you who matters in this administration's AI policy:
- Jensen Huang, who needs compute restrictions lifted so Nvidia can sell chips everywhere
- Alex Karp, whose Palantir business model depends on government AI contracts with minimal oversight
- Dario Amodei, who walked in arguing for a coordinated slowdown and walked out watching Trump reject any new federal rules
The Implication
Amodei's push for AI safety regulations just died in the White House dining room. Trump picked speed over caution, and the frontier labs that share his "go fast" philosophy now have clear sailing. Expect the gap between American and international AI governance to widen. The EU is building regulatory frameworks. China is building with state control baked in. America just decided it's building without brakes.
If you're building agent infrastructure, you now know the regulatory climate for the next three years: permissive, competition-focused, allergic to slowdowns. If you're worried about AI alignment or safety, you just lost your biggest potential ally in Washington. The development pace is about to accelerate, and the companies arguing for caution just learned they have no leverage.
Sources
Bloomberg Tech | Fortune Tech | TechCrunch AI | Business Insider Tech