Dario Amodei, chief executive of AI startup Anthropic, met US officials at the White House on Friday in what both sides described as a productive and constructive discussion, marking a notable shift in relations between the company and the US government.
According to US media reports, Amodei met with Susie Wiles, White House Chief of Staff, and US Treasury Secretary Scott Bessent. A White House spokesperson said the meeting focused on potential collaboration and shared approaches to managing the challenges of scaling artificial intelligence technologies.
The statement added that the discussions covered cooperation opportunities as well as frameworks for addressing risks linked to advanced AI systems. The White House also said it plans to hold similar engagements with other major AI companies.
The latest tone contrasts with earlier tensions, when US President Donald Trump reportedly directed federal agencies to stop using Anthropic’s technology after the company declined to grant the Pentagon unrestricted access to its Claude AI models for military use.
Following that decision, Anthropic became involved in legal disputes with the US government, including a challenge against its designation by Defence Secretary Pete Hegseth, who placed the company on a list of firms considered a “supply chain risk.” The designation is rare for a US-based company and is typically reserved for entities from countries viewed as adversaries.
Both cases remain under judicial review.
An Anthropic spokesperson said the White House meeting focused on shared priorities such as cybersecurity, maintaining US leadership in artificial intelligence, and ensuring AI safety. The company reiterated its commitment to working with the government on responsible AI development.
Separately, Anthropic recently announced a new AI model, “Mythos,” which has not been publicly released due to potential cybersecurity concerns. The company has instead shared controlled access versions with firms including Amazon, Apple, Microsoft, CrowdStrike, and Palo Alto Networks under a project internally referred to as “Glasswing.”
Anthropic said it continues to engage with US authorities regarding AI safety and national security considerations despite earlier government directives to terminate certain contracts.