Anthropic had 90 minutes to restrict Claude Fable 5 as White House feared Chinese access

On June 12, the White House ordered Anthropic to restrict access for its Mythos-class AI models, the Claude Fable 5 and Mythos 5. The government directed the company to limit access to US nationals and roll back global access. Reportedly, Anthropic was given 90 minutes to implement restrictions with an export-control order. 

According to a Semafor report, the US government had a suspicion that a group with ties to China may have obtained access to the Mythos AI model. However, it did not reveal how they gained access or which organisation was targeted. Access to Mythos by the Chinese government could come as a threat to U.S. national security, including the possibility of reverse-engineering the model via distillation.

Must read: Anthropic Claude Fable 5 ban explained: Why US government restricted access for new AI models

Anthropic Claude Fable 5 ban

After the directive, Anthropic claimed that the White House did not provide a detailed explanation beforehand, due to the reason the company flew technical executives to meet with government officials, understand the situation, and address concerns. 

In April, Anthropic released the Mythos Preview under strict restrictions. The model was accessed by only a select group of companies and government organisations, as it carried a potential threat of misuse.

Later, adviser to President Trump, David Sacks, shared an X post in which he detailed what happened before the White House imposed export controls. Sacks revealed that the White House was informed that the Fable 5 (restricted version of Mythos) could be “jailbroken.” This means that anyone can bypass an AI system’s safety rules or restrictions.

Must read: Anthropic and White House officials to hold discussions around Claude Fable 5 ban

When Anthropic CEO Dario Amodei was informed about the flaw, he allegedly said the jailbreak was not a serious risk, and it was also mentioned in the company statement. Sacks further claimed that Anthropic refused to fix the issue. However, the allegations are not yet confirmed.

Furthermore, Amazon also reported a similar jailbreak flaw, and Amazon CEO Andy Jassy reportedly communicated the issue with members of the U.S. administration. As a result, Anthropic allegedly ended up restricting access for all users, not just foreign users.

Now, Anthropic is reported to have a meeting with the White House in Washington DC on June 22, 2026.