AI startup Anthropic improves relations with the White House ahead of its IPO - Reuters

The heat in relations between the administration of US President Donald Trump and AI startup Anthropic is gradually decreasing: the softening is manifested in separate structures of the US government, Reuters sources claim. Relations warmed after Anthropic CEO Dario Amodei visited the White House in mid-April to discuss joint work with the authorities for the first time since the conflict began, the agency writes.

One of the most significant signs of warming relations between the White House and Anthropic was a symbolic but important step: the company's CEO Dario Amodei was invited to President Trump's signing of an executive order on artificial intelligence on Ma. 21, two sources told Reuters. The event was later canceled because Trump did not like some provisions of the executive order.

Trump signed the executive order on Tuesday, June 2, and Anthropic said on social media X that he expects "cooperation" from the White House in implementing the executive order.

A White House spokesperson did not respond to Reuters' request for comment on the administration's relationship with Anthropic. The press service of the startup declined to comment to the agency.

Anthropic has also held talks with National Cybersecurity Director Sean Cairncross about Mythos, an AI model capable of finding cybersecurity vulnerabilities, an insider told Reuters.

Anthropic is still challenging in court its "supply chain risk" status assigned to the company by the Pentagon. The Defense Department is "vigorously" defending itself against Anthropic's lawsuit, Franklin Turner, a lawyer specializing in government contracts, told Reuters. The broader damage to the company's business is unlikely to diminish until the dispute with the department is resolved, he said. "Any time the government signals that it is 'washing its hands of any company,' it becomes a major problem for that company," Turner explained.

Relations between the startup and the White House soured after Anthropic banned the US military from using the company's AI models for domestic surveillance of US citizens and fully autonomous weapons systems. The US Department of Defense recognized the startup as a "supply chain risk" - the first time such a status has been given to a US company. It bans US government contractors from using Anthropic's AI when working for the US military. In addition, Trump has banned federal agencies from using Anthropic products.

Anthropic is rebuilding relations with the White House in parallel with preparations for an IPO, which could result in a $1 trillion valuation, Reuters writes. Executives of large corporations are trying to strengthen relations with the White House to avoid Trump's anger, the agency explains.

Stronger ties with the administration could boost investor confidence in the short term, PitchBook senior AI analyst Harrison Rolfes told Reuters. The current rift is "more of a short-term negative," he said.

Anthropic employees met with U.S. Treasury Secretary Scott Bessent this spring to discuss Mythos and possible presidential action on AI, a U.S. official told Reuters. Those discussions helped Trump's team draft a June 2 executive order in which the president asked top developers to hand over their most advanced models for cybersecurity testing, he said.

At the same time, Anthropic employees did not participate in the Army's April 27 AI simulation of cyberattacks, Army spokesman Sean Minton told the agency. According to the Army's statement, it was attended by the heads of cyber departments of leading AI developers, including Google and OpenAI.

This article was AI-translated and verified by a human editor