Anthropic Pulls Fable 5 and Mythos 5 for All Users

As of 5:12 p.m. ET on Friday evening, Anthropic made the decision to pull Fable 5 and Mythos 5 from global availability. While other Claude models remain accessible, these two flagship models are now completely offline for every user, everywhere.

The trigger came from the U. S. government, which issued an export directive citing national security authorities. According to Anthropic’s announcement, the directive required the company to block access to both models by “any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.” The practical effect was immediate and total: all users, regardless of location, lost access.

I experienced the impact firsthand in my Claude Code session. When I tried to launch with Claude Fable 5 selected, the system responded that “Claude Fable 5 is currently unavailable” and suggested there was an issue with the model’s existence or my access permissions.

Anthropic’s statement on the matter is carefully worded. The company initially called the government’s communication a “directive” but later described it as a letter that “did not provide specific details of its national security concern.” The company’s understanding is that “the government believes it has become aware of a method of bypassing, or ‘jailbreaking’ Fable 5.”

Anthropic noted that it saw a demonstration of a “specific technique” used to identify “a small number of previously known, minor vulnerabilities.” The company characterized all the vulnerabilities as “relatively simple” and stated that “other publicly-available models are able to discover them as well without requiring a bypass.” This raises questions about whether competitor models can find flaws in Fable 5 and Mythos 5, or whether those same flaws exist elsewhere.

The company reiterated its safeguards, emphasizing that “No testers have yet been able to find a universal jailbreak.” Anthropic also acknowledged a realistic concern: “We suspect that perfect jailbreak resistance is not currently possible for any model provider. Every safeguard used in the industry is vulnerable to non-universal jailbreaks, and it is likely that universal jailbreaks will eventually be found in the future.”

According to Anthropic, the government has only provided “verbal evidence of a potential narrow, non-universal jailbreak.” This jailbreak reportedly involves instructing the model to read a “specific codebase and fix any software flaws.” The company has reviewed a report that it believes is the basis for the government’s directive and pushed back against the specific targeting of Fable 5 and Mythos 5. Anthropic validated that the level of capability displayed is not unique to these two models and is available from others, including OpenAI’s GPT-5.5. The company also noted that this capability is “used every day by the defenders who keep systems safe.”

Over the weekend, additional details emerged from credible sources. David Sacks, former White House special advisor for AI and crypto, posted on X that the administration asked Anthropic to fix the jailbreak or pull the model, and that Dario Amodei refused. Politico reported that Amazon CEO Andy Jassy flagged the problem to the White House, though The Verge noted that Amazon’s own security research led to the ban, not just a phone call. The Wall Street Journal reported that Amazon took the findings to the Commerce Department, which imposed the ban. Axios added that Commerce gave Anthropic about 90 minutes on Friday to take the models down, with the formal control letter arriving at 5:30 p.m.

The Information reported that an unnamed U. S. official said the White House is unlikely to extend the restrictions to other AI companies. Two government tweets also surfaced. Pete Hegseth, U. S. secretary of war, posted, “Three months ago, @DeptofWar kicked @AnthropicAl out of our building-forever. Every passing day proves why that was the right move.” Kirsten Davies, Chief Information Officer of the Department of War, added, “We fully support @POTUS and @SecWar in prioritizing national security. Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation.”

As of Sunday evening, Axios reported that Anthropic senior technical staff were dispatched to Washington for face-to-face discussions with the White House, transitioning from virtual conversations to real-world negotiations.

Anthropic disagrees with the directive. The company stated that “the government should have the ability to block unsafe deployments, as part of a statutory process that is transparent, fair, clear, and grounded in technical facts.” While following the directive and removing access, Anthropic argued that “the finding of a narrow potential jailbreak should not be cause for recalling a commercial model deployed to hundreds of millions of people.” The company warned that if this type of response were applied “across the industry,” it would have chilling effects on new model deployments by all frontier providers. Anthropic ended by apologizing for the disruption and expressing belief that “this is a misunderstanding,” working to restore access as soon as possible.

This situation presents a backhanded marketing win for Anthropic. On one hand, the national security implications are genuine. As AI capabilities advance rapidly, we face new challenges and questions. On the other hand, having the White House halt exports because a product is too powerful makes a compelling case that Claude Fable and Mythos are the real deal. Most software vendors spend huge marketing budgets trying to prove their products can deliver. Anthropic now has the government itself validating that concern. Nobody wants this kind of marketing when a flagship product might be shot down, but if Anthropic can turn the situation around and get Fable back into the market, it has certainly made the case for why people might pay double to use it.