Dario Amodei, co-founder and chief government officer of Anthropic, throughout an interview at Anthropic’s headquarters in San Francisco, California, US, on Thursday, April 30, 2026.
Jason Henry | Bloomberg | Getty Photos
Anthropic CEO Dario Amodei will be a part of President Donald Trump for a non-public dinner on the White Home on Sunday, CNBC confirmed, in what would be the first one-on-one assembly between the 2 leaders.
Amodei was noticeably absent from Trump’s glitzy state dinner for Chinese language President Xi Jinping on Thursday, as different tech bigwigs — OpenAI CEO Sam Altman, SpaceX CEO Elon Musk, Nvidia CEO Jensen Huang and Meta CEO Mark Zuckerberg — had been photographed rubbing elbows with authorities officers. Amodei had a scheduling battle, so Trump invited him to a non-public dinner, in accordance with a supply accustomed to the small print who was not licensed to talk publicly.
The assembly, first reported by Axios, comes at a high-stakes second for Anthropic, which is gearing up for what’s extensively anticipated to be a blockbuster IPO. The corporate’s precarious relationship with the Trump administration has been under strain for a lot of the final 12 months, so Amodei’s sit-down with the president may function a useful alternative to ease tensions.
“President Trump has been clear: America will lead the world in Tremendous Intelligence, whereas defending American customers,” a White Home official mentioned in an announcement, in response to CNBC’s inquiry about Sunday’s dinner. “We are going to drive innovation, strengthen American competitiveness, and guarantee this transformative know-how works for the American individuals.”
Amodei drew Trump’s ire earlier this month by publicly urging the AI business to slow down the pace of superior mannequin improvement and higher deal with potential risks. He printed a three-step proposal and known as for extra authorities oversight, which angered the president even because it gained the help of different business leaders like Altman and Musk.
“The one management or ‘guardrails’ that AI wants is a STRONG AND SMART (Excessive IQ!) PRESIDENT, and the united statesA. has that, in spades!,” Trump wrote in a put up on Truth Social. “The Trump Administration has stopped AI ‘individuals’ from doing unhealthy, or doubtlessly unhealthy, ‘issues,’ like Dario (Anthropic!), who’s now pretending to be a ‘excellent little angel’ – and we are going to proceed to take action!”
Amodei was already in a tough spot with the president, as a result of he supported Kamala Harris within the 2024 presidential election and was a vocal critic of Trump.
However earlier than Amodei’s name for a slowdown rocked the tech sector, Anthropic’s relationship with the Trump administration gave the impression to be on the mend.
OpenAI CEO Sam Altman (L) and Meta CEO Mark Zuckerberg (R) attend a state dinner hosted by President Donald Trump for Chinese language President Xi Jinping within the East Room of the White Home in Washington, DC, on September 24, 2026.
Brendan Smialowski | Afp | Getty Photos
Commerce Secretary Howard Lutnick, who briefly imposed export controls on two of the corporate’s most superior fashions in June, instructed Axios earlier this month that Anthropic is “again on the precise aspect.” He additionally hosted a fireside chat with Anthropic co-founder Tom Brown on the G20 Innovation Ministerial in North Carolina.
Brown, who serves as Anthropic’s chief compute officer, has taken on a way more energetic function in interfacing with the Trump administration in latest months. He reportedly led the corporate’s negotiations with the Commerce Division over its export management directive, changing Amodei in that function.
“We belief Anthropic,” Lutnick mentioned on the sidelines of the occasion.
That very same day, White Home Workplace of Science and Expertise Coverage (OSTP) Director Michael Kratsios told CNBC that the administration was “in cadence” with Anthropic. He mentioned the corporate was “open” and actively sharing fashions with the federal government as a part of Trump’s voluntary AI government order.
Anthropic has had much less success successful over officers on the Division of Protection, nevertheless.
Negotiations between the corporate and the Pentagon spiraled in February after the 2 sides did not agree on how the army may use its fashions. The DOD wished Anthropic to grant the army unfettered entry to its fashions throughout all lawful functions, whereas Anthropic wished assurance that its know-how wouldn’t be used for totally autonomous weapons or home mass surveillance.
After talks collapsed, the DOD labeled Anthropic a supply chain risk, an unprecedented transfer that meant the corporate purportedly threatened U.S. nationwide safety. The designation prevents the U.S. army from utilizing Anthropic’s fashions and blocks protection contractors from utilizing them of their work with the company.
Anthropic sued the Trump administration in an effort to reverse it is blacklisting, however the firm suffered a blow on Friday after a federal appeals court docket panel in Washington, D.C., upheld the Pentagon’s action. The corporate may nonetheless petition for a rehearing, and will additionally ask the Supreme Court docket to take the case.
Emil Michael, the DOD’s know-how chief, celebrated the ruling in a post on X.
“The hammer of justice has smashed @AnthropicAI arguments,” Michael wrote on Friday. “They’re a Provide Chain Threat to the protection industrial base serving the @DeptofWar. Warfighters will sleep higher realizing that no non-public firm will insert their opinions within the chain of command. @SecWar was proper!”
Protection Secretary Hegseth accused Anthropic of “duplicity” and argued that the corporate tried “to strong-arm america army into submission” when negotiations collapsed in February. He reaffirmed that place in a post on X on Friday.
“Confirmed: @AnthropicAI = Provide Chain Threat,” Hegseth wrote. The @DeptofWar does what is correct for the Nation and our Warriors.”
WATCH: Pres. Trump and Pres. Xi share a toast at state dinner surrounded by U.S. executives
