# US restricts Anthropic Fable 5 and Mythos access > Source: > Published: 2026-06-13 09:51:29.339167+00:00 # US restricts Anthropic Fable 5 and Mythos access The U.S. government issued an export-control directive on June 12 that suspended foreign-national access to Anthropic's two most advanced models, Fable 5 and Mythos 5, according to Anthropic's statement on its website. Commerce Secretary Howard Lutnick sent a letter to CEO Dario Amodei citing national security concerns after another company claimed it had found a way to jailbreak Mythos, per Axios. Anthropic wrote that it received the directive at **5:21pm (ET)** and that the order required the company to "abruptly disable Fable 5 and Mythos 5 for all our customers" to ensure compliance. Anthropic says it is complying while disagreeing with the rationale, calling the government's evidence a "narrow, non-universal jailbreak," and is working to restore access. Reuters and BBC report authorities cited a claimed jailbreak method that could let the models identify software vulnerabilities. Editorial analysis: this marks a notable escalation in export-control use directly targeting AI models rather than hardware or tooling. ### What happened Anthropic posted a statement on its website on June 12 saying the U.S. government issued an export-control directive that suspended access by any foreign national to its two most advanced models, Fable 5 and Mythos 5. Commerce Secretary Howard Lutnick sent a letter to CEO Dario Amodei citing national security concerns after another company claimed it had found a way to jailbreak Mythos, per Axios. The company wrote that it received the directive at **5:21pm (ET)** and that "the net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for **all** our customers to ensure compliance." Reuters and BBC report U.S. authorities cited national security concerns tied to a purported method of bypassing safeguards, described in coverage as a way to "jailbreak" the models and use them to identify software vulnerabilities. Anthropic says it is complying with the directive while disagreeing with the rationale and is working to restore access. ### Technical details Anthropic's statement says the company reviewed a demonstration of the alleged bypass and "found a small number of previously known, minor vulnerabilities," and that other publicly available models can discover those vulnerabilities without a bypass. Reporting by CNBC and Reuters notes that Fable 5 was presented as a version of Anthropic's Mythos family with strengthened safeguards and that the models had been subject to pre-release red-teaming with government and private partners, per Anthropic's account. ### Editorial analysis - technical context Observers tracking model risk note that models capable of high-fidelity code reasoning or vulnerability discovery raise distinct operational-security questions. Industry-pattern observations: researchers and vendors frequently find that increased model capability often surfaces new adversarial vectors, including jailbreak techniques that repurpose models for reconnaissance or exploit generation. These vectors drive interest from regulators and national-security agencies in ways that differ from traditional hardware-focused export controls. ### Context and significance Editorial analysis: multiple outlets frame the directive as an escalation in U.S. efforts to limit foreign access to advanced AI capability, shifting some export-control attention from chips and tooling to the models themselves, Reuters reports. Technology leaders in non-U.S. markets reacted by urging investment in sovereign AI capacity and open-source models, per regional technology reporting; those reactions reflect broader concern about dependence on foreign-controlled model providers. ### What to watch - •Regulatory follow-through: whether the U.S. Commerce Department or another agency publishes formal guidance narrowing the scope or criteria for model export controls, as reported details so far are limited. - •Vendor responses: whether other advanced-model providers change distribution controls, geofencing, or legal arrangements after this directive is reported. - •Technical mitigations: public disclosure from vendors or independent researchers on the jailbreak technique described, and whether fixes are simple rule-based patches or require architecture-level changes. ## Scoring Rationale U.S. government forcing a frontier AI lab to disable its two most advanced models globally - citing national security and export-control law - is a landmark policy action that shifts export-control frameworks from hardware and chips to model access itself. It directly and immediately affects practitioners' access strategies, vendor risk assessments, and compliance planning worldwide. Practice interview problems based on real data 1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)