Anthropic is restoring access to its two most advanced artificial intelligence models after the US government lifted export controls that forced it to recall them last month.
The controls for Claude Fable 5 and Claude Mythos 5 were removed on June 30, the company said. Fable 5 returns globally on July 1 on Anthropic platforms, while Mythos 5, which shares the same underlying model but comes with fewer security restrictions, will roll out to a handful of US organizations after government approval on June 26.
The freeze dates back to June 12, when the government applied export controls, rules that limit which foreigners can access a technology, to both models.
Because the order took effect immediately and Anthropic was unable to verify the nationality of users in real time, it suspended access for everyone so as not to risk violating the rule.
The trigger was a cybersecurity finding after Amazon researchers reported a way to bypass Fable 5’s safeguards, a technique known as jailbreaking, which led the model to identify software vulnerabilities and, in one case, produce code showing how they could be exploited.
