U.S. Orders Anthropic to Suspend Fable 5 and Mythos 5 Access for Foreign Nationals
Anthropic says it will disable access to Claude Fable 5 and Mythos 5 after a U.S. order citing national security concerns.
Intelligence analysis by GPT-5.4 Mini
Anthropic says U.S. authorities ordered it to cut foreign nationals off from its newest models, prompting the company to temporarily suspend access while it disputes the basis for the directive. The fight centers on whether a reported jailbreak shows a real national-security risk or just a narrow flaw already found in other models.
A robot helper company was told by the U.S. government to block some of its newest helpers for foreign people. The company says the problem may be small and already known, like finding one weak lock on a big house.
Analysis
What happened
Anthropic said it received a U.S. government order at 5:21 p.m. ET requiring it to suspend access to Claude Fable 5 and Mythos 5 for foreign nationals, whether they are inside or outside the United States. The company said it will "abruptly disable" access while it tries to restore service and argues the government may be acting on a misunderstanding.
The dispute
According to Anthropic, the government believes someone found a way to bypass or "jailbreak" Fable 5. The company says it reviewed that demonstration and found it exposed only a small set of previously known, minor vulnerabilities, adding that similar issues can also be found by other publicly available models without any bypass.
Anthropic also said the order does not affect its other models. It noted that Mythos 5, which shares the same underlying model but has fewer safeguards in some cybersecurity areas, remains available only to vetted cyber defenders and critical infrastructure operators.
Security context
The article says Anthropic has built guardrails and safety classifiers intended to stop misuse, including jailbreak attempts and harmful cyber-related requests. It also says the company has recently warned that Mythos-class models can turn newly disclosed bugs into working exploits in hours or minutes, which raises the stakes around both offensive and defensive use.
Anthropic argues that a narrow jailbreak should not be used as the basis for recalling a commercially deployed model, and it wants the statutory process to be "transparent, fair, clear, and grounded in technical facts." The company also said no universal jailbreak has been shown against its latest models so far.
The story comes after earlier U.S. defense scrutiny, with the Department of Defense labeling Anthropic a supply chain risk and the company filing lawsuits to fight that designation.
Key points
- Anthropic says the U.S. government ordered it to suspend foreign-national access to Fable 5 and Mythos 5.
- The company says it believes the directive is based on a narrow jailbreak, not a universal model failure.
- Anthropic claims the reported flaw exposed only small, previously known vulnerabilities.
- Mythos 5 remains available only to vetted defenders and critical infrastructure operators.
- The company argues the process should be transparent and grounded in technical facts.
If Anthropic is correct, the review could lead to a more precise, technically grounded policy instead of a broad shutdown. The company may also use the incident to strengthen safeguards and clarify which users can safely access its most capable models.
If the government sees the jailbreak risk as real, access restrictions could expand and slow deployment of advanced AI tools. The episode could also deepen tensions between AI vendors and regulators over who decides when a model is too risky to remain widely available.
