The 15-Day Ban Is Over
It's official: Anthropic's Claude Mythos 5 is back — but not for everyone.
On June 27, 2026, the US government lifted its block on Anthropic's most advanced AI model, allowing the company to release it to over 100 trusted institutions, including major corporations and government agencies. The move ends a dramatic 15-day standoff that began on June 12 when the Trump administration issued an emergency order restricting both Mythos 5 and Fable 5 over national security concerns.
What Changed?
The partial unblock comes after Anthropic demonstrated rigorous new safety protocols, including:
- Real-time monitoring of model outputs for sensitive queries
- Institutional access controls restricting who can deploy the model
- Audit logging with government oversight
Unlike the broad consumer block on Fable 5 (which remains restricted), Mythos 5 is now available to "trusted partners" — think Fortune 500 companies, defense contractors, and federal agencies.
The Fallout
The decision has also impacted OpenAI. Hours after the Mythos announcement, OpenAI confirmed it would stagger the rollout of GPT-5.6 Sol following a separate White House request — a move CEO Sam Altman described as "prudent" but critics call a dangerous precedent for government AI control.
Meanwhile, Anthropic's stock surged 14% in after-hours trading. The message from Washington is clear: powerful AI isn't banned — it's just gatekept.
What's Next
The Mythos 5 unblock sets a major precedent. With the EU AI Act deadline just 36 days away (August 2), the US is sending a signal that it prefers institutional licensing over blanket bans. Expect every frontier lab to be studying this playbook closely.
The age of government-approved AI access has begun.
Top comments (0)