The Trump administration has ordered Anthropic to remove its most advanced AI models, including Claude Mythos and Fable 5, from public access. This decision follows a directive issued nearly a week ago, which has left the company in a dispute with the White House over how to safely reintroduce these models. Anthropic maintains it followed all procedures, but the administration argues the company failed to ensure safe deployment of frontier AI technology.
The situation highlights the lack of clear regulations for AI development in the U.S., with officials expressing concerns over potential national security risks. A former White House technology official noted that the administration’s anti-regulatory stance has led to an ad-hoc approach to managing AI risks, creating uncertainty for the industry. The White House has repeatedly opposed efforts to impose guardrails on AI, arguing such measures could harm U.S. innovation and competitiveness.
The dispute remains opaque, with the U.S. government not providing specific details on what Anthropic did wrong. A post on X from White House technology adviser David Sacks outlines the general situation, but no clear guidelines have been issued. This lack of transparency has raised concerns about how the administration will handle future AI regulatory challenges. Anthropic’s models, which the company claims have accelerated its research, are now inaccessible to employees, customers, and partners, including major tech firms like Apple and Meta.
Source: wired
