THE VERTEX.
Back to home
TECHNOLOGY28 June 2026

Trump Administration Reopens Access to Anthropic’s Flagship Model for U.S. Stakeholders

The White House has restored access to Anthropic’s most advanced AI model, Mythos, for a select group of U.S. companies and government agencies after weeks of negotiation.

La
La Rédaction
The Vertex
5 min read
Trump Administration Reopens Access to Anthropic’s Flagship Model for U.S. Stakeholders
Source: www.wired.com
In a quietly orchestrated move that underscores the growing intersection of federal policy and frontier artificial intelligence, the White House announced last week that it has restored access to Anthropic’s latest model, codenamed Mythos, for a curated cohort of American enterprises and federal agencies. The reversal follows weeks of intensive negotiation between senior administration officials and Anthropic executives, who argued that selective deployment would enable critical sectors—including defense, healthcare, and energy—to evaluate the model’s capabilities under strict oversight. By limiting access to vetted entities, the administration seeks to balance rapid technological advancement with national security considerations, a tension that has defined U.S. AI strategy since the release of large language models in 2023. This approach also allows the government to monitor potential misuse while fostering responsible innovation across the private sector. This development fits into a broader narrative of the United States attempting to steer the AI race without stifling innovation. While the European Union tightens its AI Act and China accelerates its own model releases, Washington’s selective openness signals a willingness to harness commercial talent while maintaining a strategic choke‑point over the most powerful systems. The move also reflects rising pressure on tech firms to demonstrate responsible stewardship of dual‑use technologies. Looking ahead, the limited rollout may serve as a template for future regulatory frameworks, where access is granted contingent on auditability and alignment with policy objectives. If successful, it could accelerate the integration of advanced AI into critical infrastructure, but it also raises questions about equity, market concentration, and the potential for misuse by actors outside the approved circle. The ultimate impact will depend on how rigorously the administration enforces the conditions attached to Mythos access.