Anthropic CEO Set for First One-on-One Dinner With Trump
Anthropic chief executive Dario Amodei was scheduled to have dinner with President Donald Trump at the White House on the evening of September 27, 2026.
It would be the first one-on-one meeting between Amodei and Trump. Axios first reported the plans on September 27, and TechCrunch confirmed them with a source familiar with his plans.
The invitation came after Amodei was absent from Trump's state dinner for Chinese President Xi Jinping. Executives from OpenAI, Google and Meta were on the guest list for that dinner, according to Fortune India. The OpenAI CEO and the Nvidia CEO were also set to attend. Amodei was not invited.
Forbes put the combined net worth of tech leaders attending the Trump-Xi dinner at $2.2 trillion and put Amodei's own net worth at $15.5 billion. Amodei had urged top AI firms to slow down AI model development. He had also written that China taking the lead in the AI race would pose a grave danger to the US and the world, as noted by Firstpost.
On policy, Amodei refused to allow the Department of War to use Claude, Anthropic's AI assistant, for mass domestic surveillance. He had met earlier with White House Chief of Staff Susie Wiles to discuss Anthropic's Mythos AI model, according to CNBC. That meeting took place in April.
Anthropic shipped three updates in September. It introduced Claude Fable 5.1 and Claude Mythos 5.1, two versions of its AI system tuned for different uses, on September 1, 2026. It published "Detecting and countering misuse of AI: September 2026" on September 10. It introduced Claude Opus 5.5 on September 22. The company said Opus 5.5 performs at the level of Claude Fable 5.1 on most work and costs 40% less to run than Opus 5. That running cost is the computing power needed to answer user requests.
Amodei was parodied on the season premiere of Saturday Night Live on the weekend of September 27, 2026. It is uncommon for a lab CEO to appear in late-night comedy and at a White House dinner in the same weekend.
The broader context here is a negotiation over access and terms, not only a calendar invite. Leading AI labs need computing power, distribution to users and clear rules, while Washington is managing AI competition and safeguards for deployment. A direct conversation lets both sides test where commercial plans and national priorities line up and where they do not.
In my view, the practical item to watch is narrow. Efficiency gains in Opus 5.5 and the parallel Fable and Mythos lines give Anthropic more options to discuss, from lower-cost business use to AI that can handle multi-step tasks on its own. A one-on-one dinner does not resolve surveillance limits, pacing debates or China policy. It does create a channel where those issues can be addressed with specific technical limits on the table, which is how durable agreements in this industry have usually started. The long arc still favors more capable and cheaper models reaching more users, if the guardrails hold.
