Anthropic Sidelines Amodei in White House AI Talks

Anthropic has reportedly pulled CEO Dario Amodei from its most sensitive White House AI talks, handing the file to cofounder Tom Brown as negotiations with the Trump administration heat up over security and export rules.

Still CEO, but out of the room

Dario Amodei reportedly remains firmly in charge of Anthropic, yet he has been quietly stepped back from the company's most delicate dealings with Washington. According to a report from Crypto Briefing on June 24, 2026, cofounder Tom Brown has taken over leading Anthropic's discussions with Trump administration officials on AI security policy, leaving the CEO with a quieter seat at the table on this particular front.

The shift is notable for a company that has built so much of its public identity on engaging seriously with regulators. Brown is said to be meeting figures including Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross. The report suggests the handoff followed friction over how aggressively Anthropic should accommodate the government's demands, with one White House official having reportedly gone so far as to label Amodei a "weirdo."

What the negotiations are really about

The substance of the talks goes well beyond personality. Per Crypto Briefing, the discussions center on two intertwined questions: export controls governing advanced AI, and the thornier problem of how to actually measure when an AI system's safety has failed.

That second issue reportedly has roots inside Anthropic itself. The outlet says a "jailbreak incident" involving the company's Mythos and Fable models touched off an internal disagreement between Brown and Amodei over the correct compliance approach, a dispute that appears to have spilled into how the company now presents itself to officials.

Key threads from the report:

Talks focus on creating benchmarks to standardize how AI "jailbreak" incidents are measured.
Those benchmarks could ultimately feed into a broader export-control framework.
Brown has reportedly been steering the discussions since around mid-June.
A tokenized representation of Anthropic stock was trading near $740 on a secondary platform, underscoring intense investor interest.

The push toward shared benchmarks reflects a wider reality in AI policy, where regulators and companies alike have struggled to agree on objective measures of safety. Without an agreed yardstick for what counts as a failure, rules around what models can be built, deployed or exported remain difficult to write and harder to enforce.

Why the reshuffle matters

The reported personnel switch is about far more than who attends which meeting. It highlights how decisively access to policymakers now shapes the fortunes of the AI industry. The ability to influence emerging rules can determine which models a company is permitted to deploy at home and ship abroad, making the relationship with Washington a strategic asset in its own right.

That is what makes the optics so pointed for Anthropic specifically. The company has marketed itself relentlessly on safety-first messaging, positioning its founder as a thoughtful voice on the risks of powerful AI. Yet the report suggests that same founder-CEO struggled to maintain a productive rapport with key administration figures, prompting the company to route its most consequential conversations through a cofounder seen as easier to work with.

Handing the file to Brown signals just how high the stakes have climbed. With Washington weighing rules that could reshape the entire competitive landscape, Anthropic appears to be prioritizing a smooth working relationship with regulators over insisting that its chief executive personally carry every negotiation. Whether that pragmatism pays off, or reads as the company blinking under pressure, may depend on what framework finally emerges from the talks.