Anthropic blocks access to Fable 5 and Mythos 5 after US government order barring all foreign nationals
Anthropic has cut off access to its two most advanced AI models, Fable 5 and Mythos 5, for all non-US persons after the US government ordered the move citing national security. The restriction applies even to foreign nationals inside the United States and on Anthropic's own staff.
The block order
Anthropic disclosed on 13 June 2026 that it had been instructed by US government agencies to immediately block access to its Fable 5 and Mythos 5 models for all foreign nationals. The order, issued on national security grounds, covers foreigners anywhere in the world, including those physically present in the United States and employees of the company itself. Anthropic said it was forced to cut off access for everyone at short notice while it worked through the implications.
The firm stated it had received only partial information from the government about the alleged risk. Anthropic reviewed what it believes was the triggering report and concluded the identified issue was a limited capability that allowed the AI to inspect specific program code and fix errors.
Why Fable 5 was blocked
Fable 5 was released earlier this week, on Tuesday 9 June, as a consumer-facing version of the far more powerful Claude Mythos system. While Fable 5 inherits Mythos's technical architecture, its cybersecurity and biotechnology capabilities were deliberately blocked before the public launch. The model is designed to refuse questions touching those fields and redirects users to the older Claude Opus 4.8.
Anthropic insists that Fable 5's safety measures were extensively tested and that other providers' models, such as OpenAI's GPT-5.5, possess similar code-review capabilities. The company expressed disagreement that software serving hundreds of millions of users should be blocked on the basis of the reported finding.
- Anthropic clashes with the Pentagon over autonomous weapons; labeled a supply-chain risk.
- Anthropic announces Claude Mythos but does not release it publicly, citing safety concerns.
- Fable 5 is released as a safer consumer version of Mythos, with cybersecurity and biotech abilities blocked.
- US government orders Anthropic to block all foreign access to Fable 5 and Mythos 5.
Anthropic's dispute with Washington
The order deepens a running conflict between Anthropic and the US government. In March 2026, the company clashed with the Pentagon after refusing to allow its AI to be used in autonomous weapons systems or for mass surveillance inside the United States. The Pentagon subsequently designated Anthropic a supply-chain risk, a move that can severely hamper government adoption of its software. Anthropic is challenging that designation in court.
Ironically, CEO Dario Amodei had only days earlier advocated for government authority to block potentially dangerous AI software. Anthropic now stresses that any such action must follow clear, transparent procedures grounded in technical facts, which it says did not happen here.
What Mythos can do
The full Claude Mythos model, which remains restricted to US government agencies and selected corporate partners, earned worldwide attention in April 2026 for its ability to autonomously find and exploit software vulnerabilities, some of which had lain undetected for decades. That capability has been used to patch security holes but also raised fears that in the wrong hands it could become a dangerous cyberweapon. Germany's Federal Office for Information Security (BSI) sounded an alarm at the time of the model's announcement.
Broader context
The block comes just as Anthropic was preparing for a planned stock market listing and days after the company called on the world to end the AI arms race and agree to a global development pause. The sudden restriction throws both the IPO timeline and the company's international cooperation into uncertainty.
