In brief
Anthropic abruptly paused distribution of its Claude Fable 5 model after U.S. officials flagged potential security risks. The company said it had to disable Fable 5 and its companion Mythos 5 for all users, citing an order to restrict foreign nationals. While the firm contends the government did not pinpoint specific threats, it acknowledges a possible jailbreak method. In parallel, Anthropic is embroiled in a lawsuit with the Trump‑era administration and the Pentagon over AI tool use by government agencies, a case that a judge has so far ruled unenforced, allowing continued use while legal battles continue. The incident underscores growing scrutiny of high‑power AI systems and their safety safeguards.
Anthropic has temporarily suspended its newly released AI model, Claude Fable 5, after U.S. authorities expressed security concerns. The tech company said it must abort the tool for all users to stay compliant with an order that restricts foreign nationals from using the model.
In a statement on its website, Anthropic explained that the net effect of the order is that it must suddenly disable Fable 5 and its counterpart Mythos 5 for every customer, stating the tools were deemed “too powerful” for public release.
The firm is also in a legal battle with the Trump administration. The U.S. Department of Commerce has designated Anthropic a “supply‑chain risk,” a designation typically reserved for companies from rival nations. A U.S. judge has ruled that the Pentagon’s directive to stop using the company’s AI cannot be enforced, allowing agencies to continue the partnership while the lawsuit runs.
Anthropic’s flagship model, Claude Fable 5, is part of its Claude Mythos line, directly competing with OpenAI’s ChatGPT and Google’s Gemini. Officials feared a jailbreak— a technique to subvert the software’s sanity checks— could expose vulnerabilities. Although Anthropic found only minor, publicly detectable weaknesses, the model’s ability to exploit system flaws prompted the pre‑release to limited organisations for safety checks.
Critics say Anthropic’s claims of the tool being “too powerful” were largely marketing hype, yet the company insists its safeguards are robust. The incident highlights the growing tension between rapid AI innovation and national security oversight.




