US lifts export controls on Anthropic’s Claude Fable 5
The U.S. government has lifted export controls on Anthropic's Claude Fable 5 and Mythos 5 models, allowing them to be redeployed with new safeguards.
The update
Anthropic is restoring public access to its most powerful AI models, Claude Fable 5 and Mythos 5, after the U.S. government lifted export controls that forced a suspension last month. The controls were originally applied on June 12 following a report that researchers had bypassed Fable 5’s safeguards. The government cleared the models on June 30, and Fable 5 is now returning globally on July 1. Mythos 5, a less restricted variant, is being reintroduced to approved U.S. organizations.
Why it matters
This reversal marks a shift in U.S. AI security policy, allowing advanced frontier models to be deployed with updated safeguards. The episode highlights the White House’s focus on maintaining American leadership in AI while managing cybersecurity risks. The restrictions also raised concerns about state control over frontier technology, prompting closer collaboration between Anthropic and U.S. agencies on model testing and threat sharing.
What to watch
Anthropic is deploying new classifiers to block a broader range of cybersecurity-related tasks. The company is also drafting a consensus framework with partners in Project Glasswing to assess jailbreak risks. It remains unclear how this reversal will impact other AI companies facing similar export restrictions or how the new safeguards will perform under real-world use.
Sources
- Cointelegraph — Context on the suspension, government clearance, and new classifiers.
- CoinDesk — Details on the timeline, Mythos 5 restrictions, and the jailbreak trigger.
