The US government issued an export control directive forcing Anthropic to disable its Fable 5 and Mythos 5 AI models for all users worldwide.

Anthropic said in a statement published June 12, 2026, that it received an export control directive from the US government at 5:21 p.m. ET, citing national security authorities and ordering the suspension of all access to Fable 5 and Mythos 5 by any foreign national, whether located inside or outside the United States, including the company's own foreign-national employees.
Anthropic said the directive's practical effect required it to disable both models for every customer to ensure compliance, while access to all other Anthropic models remained unaffected. According to NBC News reporter Jared Perlo, Claude's landing page displayed a notice stating that "Fable 5 is temporarily unavailable" following the shutdown.
Anthropic said the directive did not provide specific details of the government's national security concern, but the company's understanding was that officials had become aware of a method for bypassing, or "jailbreaking," Fable 5's safeguards. The directive was reportedly sent from Commerce Secretary Howard Lutnick and drafted with assistance from officials at the Commerce Department's Bureau of Industry and Security.
Anthropic said it reviewed a demonstration of the technique and found it had been used to identify a small number of previously known, minor vulnerabilities, which the company said other publicly available models can also discover without requiring any bypass. The company stated that the government had so far provided only verbal evidence of a narrow, non-universal jailbreak that involved asking the model to read a specific codebase and fix software flaws, and that this capability is also available from other models, including OpenAI's GPT-5.5. Anthropic said it disagreed that a narrow potential jailbreak should justify recalling a commercial model deployed to hundreds of millions of users, and described the action as based on what it called a misunderstanding.
In its statement, Anthropic outlined its safeguard strategy for Fable 5, saying the model had been red-teamed for thousands of hours with the US government, the UK AI Safety Institute, and third-party organizations prior to launch, and that no tester had found a universal jailbreak. The company said it had adopted a 30-day data retention policy for Mythos-class models specifically to support jailbreak research and mitigation.
Anthropic released Fable 5 and Mythos 5 on Tuesday, June 9, 2026, describing them as the most capable AI systems the company had produced. Fable 5 was made publicly available with stronger restrictions on cybersecurity- and biology-related queries, while Mythos 5 was released without those safeguards to a select group of partners, including cybersecurity and infrastructure companies. The Financial Times had previously disclosed that the National Security Agency was using Mythos to conduct offensive cyberattacks.
Reports suggest that the suspension follows a dispute earlier in 2026, when President Donald Trump and Defense Secretary Pete Hegseth moved in February to restrict Anthropic's products from federal agencies after the company sought additional guardrails on Pentagon use of its technology, with Trump criticizing the company in a post on Truth Social.
Anthropic subsequently sued the administration over the matter, and that a federal judge in California ruled in Anthropic's favor, though the case remains active in a Washington, D.C. federal court.
Anthropic said it believes the government should retain the authority to block unsafe deployments through a transparent, statutory process grounded in technical facts, but stated that the directive issued June 12 did not adhere to those principles. The company said it was complying with the order while working to restore access to Fable 5 and Mythos 5.
Stay informed on the fastest growing technology.
Disclaimer: The content on this page and all pages are for informational purposes only. We use AI to develop and improve our content — we love to use the tools we promote.
Course creators can promote their courses with us and AI apps Founders can get featured mentions on our website, send us an email.
Simplify AI use for the masses, enable anyone to leverage artificial intelligence for problem solving, building products and services that improves lives, creates wealth and advances economies.
A small group of researchers, educators and builders across AI, finance, media, digital assets and general technology.
If we have a shot at making life better, we owe it to ourselves to take it. Artificial intelligence (AI) brings us closer to abundance in health and wealth and we're committed to playing a role in bringing the use of this technology to the masses.
We aim to promote the use of AI as much as we can. In addition to courses, we will publish free prompts, guides and news, with the help of AI in research and content optimization.
We use cookies and other software to monitor and understand our web traffic to provide relevant contents, protection and promotions. To learn how our ad partners use your data, send us an email.
© newvon | all rights reserved | sitemap

