Weeks after the U.S. government imposed control over frontier AI models of Anthropic citing national security concerns, the AI company announced Wednesday that the temporary ban on Claude Fable 5 has been lifted.
In a post on X, Anthropic announced that Claude Fable 5, a “safer” version of its frontier AI model Claude Mythos, will return to users worldwide. Anthropic also announced that it has restored access to Mythos 5 for a set of U.S. organisations, as per the directives of the government.
On June 12, Anthropic had revoked access to Claude Fable 5 and Claude Mythos to all users including its own employees after the U.S. government had put export control over the models, barring its use for all non-Americans.
“Fable 5 will be available starting tomorrow, Wednesday, July 1, to users globally on the Claude Platform, Claude.ai, Claude Code, and Claude Cowork. For Pro, Max, Team, and select Enterprise plans, Fable 5 will be included for up to 50% of weekly usage limits through July 7, after which it will be available via usage credits. We will re-enable access on AWS, Google Cloud, and Microsoft Foundry as quickly as possible,” read a statement from Anthropic.
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
According to David Sacks, the export control on Claude Mythos and Claude Fable 5 was put after researchers were able to “jailbreak” the AI models, thereby proving its safeguards could be breached for nefarious purposes.
Incidentally, the researchers were given access to Claude Mythos model under Project Glasswing of Anthropic– a shared collaborative project announced by the company in which selected organizations were given access to the Mythos model.
“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 and debugging will fall back to Opus 4.8. We’ll continue to refine these classifiers over the coming weeks to reduce false positives and better distinguish genuine misuse from legitimate requests,” read a statement by Anthropic.
Anthropic also informed that they are willing to “coordinate with the government to expand access to the broader set of domestic and international partners in the Glasswing program.”
As Anthropic announced redeploying of Claude Fable model for users worldwide, many experts say that it would be preemptive to take it as a welcome gesture and they would rather wait and observe.
Also Read: David Sacks Reveals Why White House Banned Claude Fable for Non-Americans
Speaking to AI FrontPage, Dinesh Jotwani, Supreme Court senior advocate from India, said, “We will take it as a positive news but it has come after a lot of deliberations and ifs and buts. Even after the ban, Indian government had sent a representation to the US to seek uplifting of the restriction for our agencies. Even two days ago, Anthropic announced that it will give access to Mythos model to selected group of US organizations. But I would say that it is too early and we would like to wait and watch how things unfold.”
Regarding the need for an international regulatory framework on frontier AI, Jotwani said, “Already there are talks being held on a global safety framework on AI models but we are not able to reach a consensus due to own vested interests of countries when it comes to artificial intelligence.”
Timeline of Events: Claude Mythos Preview to Claude Fable
In late April this year, Anthropic, a U.S. based AI startup, had unveiled a preview version of its frontier AI model Claude Mythos, claiming it posed cybersecurity risks due to its advanced capabilities in coding.
Anthropic announced that they will not release the Claude Mythos version for public fearing misuse of the model by non-state actors. They claimed that their frontier AI model had posed unprecedented cybersecurity risks owing to its advanced capabilities and its CEO had even demanded for a global pause on frontier AI model—something that critics like David Sacks, co-chair of the President’s Council of Advisors on Science and Technology, had termed as a “fear-mongering” market stunt.
Also Read: Anthropic Has a History of Scare Tactics: David Sacks on Claude Mythos
It instead launched Project Glasswing and invited major tech giants like Amazon, Google, Cloudflare, Oracle, Meta among others to try Claude Mythos Preview model and test vulnerabilities of their digital infrastructure. Anthropic also announced that it is willing to share the Claude Mythos model with other countries like the U.K and India.
Later, Anthropic released Claude Fable, a version of Claude Mythos, with safeguards in place to prevent its misuse. However, it was during this Project Glasswing that researchers breached the safeguards through jailbreaking of the AI model.
Following the tip by researchers, the U.S. government imposed export control over Fable and Mythos models to all non-Americans, including foreign employees of Anthropic, citing national security.
The move was criticized by several global leaders including French President Emmanuel Macron, Canadian PM Mark Henry, U.K. AI Minister Kanishka Narayan among others. The “kill switch” move on Anthropic’s AI models also made several countries notice the importance of sovereign AI.
Also Read: California, Anthropic Partner to Offer Claude to State Agencies at 50% Discount
