Anthropic restores AI models Fable, Mythos after the U.S. lifts export controls
CoinDesk 2026-07-01 06:54:13
Context: Anthropic, a tech company, has restored access to its advanced AI models, Fable 5 and Mythos 5, after the U.S. government lifted export controls that were imposed due to a cybersecurity finding. The controls were triggered by a jailbreak discovered by Amazon researchers that allowed Fable 5 to generate exploit code. The episode has led to closer collaboration between Anthropic and U.S. agencies on model testing, threat sharing, and standards for rating jailbreak risks.
Key Facts
- The U.S. government lifted export controls on Anthropic's Claude Fable 5 and Claude Mythos 5 AI models on June 30, allowing the company to restore access to the models.
- Fable 5, which was suspended on June 12, is being reintroduced globally on July 1 across Anthropic's platforms, while Mythos 5 is being restored to a set of approved U.S. organizations after government review on June 26.
- The export controls were triggered by a cybersecurity finding after Amazon researchers reported a way to bypass Fable 5's safeguards, a technique known as a jailbreak, which prompted the model to identify software vulnerabilities and produce code showing how one could be exploited.
- Anthropic is deepening its collaboration with the U.S. government, including giving designated agencies early access to frontier models and their safeguards before public release, sharing threat intelligence, and working toward a common security standard across AI developers.
- Anthropic is drafting a framework with Amazon, Microsoft, and Google for scoring how dangerous a given jailbreak is, as part of a June 2 executive order on AI security that aims to tighten the link between frontier AI releases and government review.