Why the US first blocked then restored Anthropic’s most powerful AI models has become a key question after Washington reversed export controls on Fable 5 and Mythos 5, for users around the world, the company said on Tuesday.
The restrictions had applied globally, including in India, after Washington ordered Anthropic on June 12 to deny access to foreign nationals on national-security grounds. Because the company said it could not reliably separate foreign nationals from other users in real time, it switched the models off for everyone.
Also Read | Europe heatwave claims over 1,300 lives as temperatures soar across continent
Claude Fable 5 will be available again globally tomorrow.
After a series of productive conversations with the US government, we’re redeploying the model with a new set of classifiers to target and block more cybersecurity tasks. In the near term, some routine tasks like coding…— Anthropic (@AnthropicAI) July 1, 2026
A curious turn in the AI tale
Fable 5, released in early June, is a locked-down public version of Mythos 5, which Anthropic had kept from broad release because of its unusually strong ability to identify software vulnerabilities that hackers could exploit.
Mythos 5 had been made available only to a limited set of companies and government entities under Anthropic’s Project Glasswing, a program designed to help trusted users uncover dangerous uses and weaknesses in their own systems.
Anthropic restored limited access to Mythos 5 for a small group of trusted US organizations on June 26 before the Commerce Department removed the curbs entirely on June 30.
STORY | Anthropic restores access to Fable 5, Mythos 5 AI models as US lifts export controls
Artificial intelligence major Anthropic has restored access to its advanced AI models, Fable 5 and Mythos 5, after the US government lifted export controls that abruptly halted access… pic.twitter.com/vzyiHooTtA— Press Trust of India (@PTI_News) July 1, 2026
The safeguard strikes back
According to Reuters, the export-control order followed concerns after Amazon researchers found a way to bypass Fable 5’s safeguards, allowing the model to identify software vulnerabilities or surface other illegal information.
Anthropic said it has since added a new safeguard that blocks that behavior and routes blocked requests to its older Opus 4.8 model. The company also said it was deepening cooperation with the US government, while warning that making any AI model fully resistant to jailbreaks may not be possible.
A precedent begins to emerge
The episode has broader implications for how Washington may regulate advanced AI systems.
The controls marked a shift from the US tradition of using export rules mainly on hardware to directly intervening in access to finished software products.
Anthropic also said designated government partners will receive expanded early access to its models, while a letter from Commerce Secretary Howard Lutnick said the department could revisit the decision if circumstances change or the company fails to meet its commitments.
Anthropic’s dispute with the Trump administration has already spilled into the courts and remains unresolved.
Also Read | What is a SinkEx drill and why do navies sink their own warships?
FAQs
Q1: Why did the US restrict Anthropic’s AI models?
Ans: The US imposed temporary restrictions over national security concerns after researchers demonstrated a way to bypass the models’ safety safeguards.
Q2: Why has the US restored access to Anthropic’s AI models?
Ans: The restrictions were lifted after Anthropic introduced additional safety measures and the US Commerce Department reversed the export controls.
