Trump AI Access Controls
Government approval shapes limited rollout of new AI models to select U.S. users.
Main Story
Center-RightThe Trump administration partially lifted restrictions on Anthropic’s Claude Mythos 5, allowing the company to release the powerful AI model to more than 100 trusted U.S. companies, institutions and federal agencies after a two-week block over national security and cybersecurity concerns. Commerce Secretary Howard Lutnick told Anthropic the government was confident in the company’s guardrails, marking a de-escalation in a confrontation that had forced Anthropic to pull Mythos 5 and Fable 5 from broader access. The decision still leaves Washington in an unusually direct role, determining which customers can use frontier AI systems and under what conditions. Reports also suggested Anthropic’s Fable 5 could return within days, underscoring that the restrictions may be easing but have not disappeared.
Coverage Angles
OpenAI Rollout Limits
BalancedOpenAI limited the release of its GPT-5.6 lineup — including Sol, Terra and Luna — to a small group of trusted partners after a Trump administration request for cybersecurity review. The company said it supports broad access and warned that government preapproval of model users should not become the long-term default.


