The US government has lifted export controls on Anthropic's most advanced AI models, Claude Fable 5 and Mythos 5, clearing the company to restore worldwide access after a roughly two-week suspension.
According to multiple reports, including CoinDesk and The Hacker News, the restrictions were tied to a jailbreak and national security concerns. The Decoder describes it as a two-week government ban. The US Department of Commerce lifted the controls following what one report calls a two-week collaborative review, and MSN reports the announcement came shortly after Commerce Secretary Howard Lutnick said the government had completed its review. Forbes frames the decision as coming from the Trump administration.
Anthropic said access would be restored on a global rollout beginning July 1. In a statement quoted by MSN, the company said, "We're grateful to our users for their patience and to everyone who worked with us on redeploying the models." As part of the arrangement, several outlets report Anthropic will give the US government access to the models and must cooperate with it going forward.
The restored models are not fully at capacity yet. Per Anthropic's own post (via Techmeme), "some routine tasks like coding and debugging" on Fable 5 "will fall back to Opus 4.8" in "the near term" as it works to "reduce false positives." Yahoo Tech reports Anthropic's testing showed Fable 5 wasn't uniquely risky.
Not everyone sees the episode favorably. In a thread flagged by Techmeme, security researcher Alex Stamos called the now-lifted controls an "own goal" that created uncertainty and could push service providers toward Chinese AI models.
Why it matters: the case shows how quickly the US government can now switch access to frontier AI on and off — and how those decisions ripple across global users and competition.