Anthropic CEO Heads to the White House for a Break-the-Ice Meeting: Meets with the Chief of Staff and Bessent to Discuss Mythos

According to an exclusive story from The Wall Street Journal on April 22, Anthropic CEO Dario Amodei visited the White House for a closed-door meeting on April 17. Attendees included White House chief of staff Susie Wiles and Treasury Secretary Scott Bessent. This was the highest-level conversation between Anthropic and the Pentagon since their public standoff, after which the White House said the meeting was “constructive.” The market interpreted it as a sign that relations between the two sides may be thawing.

Meeting focus: Mythos model and national security boundaries

According to reports from WSJ and CNN, the meeting centered on the “responsible deployment” issue surrounding Anthropic’s latest model, Mythos. Bessent had previously privately warned financial industry executives about the potential cybersecurity risks Mythos could pose, while the White House said it was closely monitoring the threat the model could present to critical infrastructure. Anthropic, for its part, hoped the face-to-face meeting would clarify the red line between “national security use” and a company’s acceptable use policy.

The mix of officials in attendance reflected the seriousness of the dialogue between the two sides: Susie Wiles is the White House chief of staff, directly representing Trump’s decision team; Bessent is the main spokesperson currently for AI-related financial risk issues. The fact that both attended in the same setting indicates that Anthropic’s issue has been treated as a “White House-level” matter rather than a concern confined to a single agency.

Background: The Pentagon standoff and Trump’s “Who?” remarks

Anthropic and the White House have been at odds for months, with the core disagreement being that the Department of Defense asked Anthropic to allow Claude technology for “all lawful uses.” Anthropic refused and argued for retaining discretion under its own acceptable use policy. Defense Secretary Pete Hegseth therefore labeled Anthropic a “security risk,” and Trump instructed federal agencies to cut off cooperation with Anthropic.

Afterward, when Trump was asked by reporters on the runway in Phoenix about Amodei’s visit to the White House, he responded with “Who?” and said “no idea,” which was interpreted as Trump not being aware of the meeting or intentionally keeping distance. This echoes the narrative of an Axios exclusive on April 20 that revealed the NSA used Mythos while bypassing the Pentagon blacklist—suggesting that within the federal government, attitudes toward Anthropic are evidently inconsistent.

Lobbying and dialogue on parallel tracks

The meeting also underscores two moves Anthropic has made recently: first, on April 13, it hired Ballard Partners, which is linked to Trump, as a lobbying firm; second, Q1 2026 lobbying spending reached $1.6 million, a record high. Amodei’s meeting this time can be seen as the official launch of a “dialogue track,” complementing the “lobbying track” in an attempt to achieve a policy breakthrough on both fronts.

Citing sources, the WSJ reported that the two sides did not reach any concrete agreement during the meeting, but the White House is willing to keep talking, and Anthropic is also willing to tweak parts of its acceptable use policy to address national security concerns. Further developments are expected before the Mythos rollout in early May.

Industry significance: An AI company handles a conflict with the Trump government head-on for the first time

Anthropic is currently the first AI flagship to enter a public standoff with the White House explicitly because it refused to grant the Department of Defense unlimited authorization. Other major AI companies, such as OpenAI and Google, tend to address government requests with a “compliance first” posture, but Anthropic’s insistence has forced the industry for the first time to face the triangular decision of “model capabilities vs national security vs corporate values.” The subsequent developments from this meeting are expected to become a reference case for other AI companies in similar situations.

This article, “Anthropic CEO breaks the ice with the White House: discusses Mythos with the chief of staff and Bessent,” first appeared on 链新闻 ABMedia.