Anthropic export controls AI regulation national security Fable 5

US Government Forces Anthropic to Pull Its Two Most Powerful Models Worldwide Over Foreign National Access Concerns

17 June 2026

The US government issued an export control directive ordering Anthropic to block access to its two most advanced AI models, Fable 5 and Mythos 5, for all foreign nationals, citing national security concerns. Because compliance would be impossible without disabling the models entirely, Anthropic suspended global access to both, just days after rolling out Fable 5 for free to millions of users. Anthropic disputes the justification, stating the government's evidence amounts to a narrow, already-known jailbreak, and argues that applying this standard industry-wide would effectively halt all new frontier model deployments.

Anthropic has taken its two most capable models offline globally after the US government issued an export control directive on June 12 ordering the company to block all foreign nationals from accessing them. That means Fable 5 and Mythos 5 are now unavailable to everyone, everywhere, because the practicalities of compliance apparently leave no other option.

The directive landed at 5:21pm ET and cites national security authorities. It covers foreign nationals both inside and outside the US, which notably includes Anthropic's own foreign-national staff. When your compliance burden extends to your own workforce, blanket suspension becomes the path of least resistance.

The timing stings. Anthropic had only started rolling out Fable 5 on June 9, offering it free to Pro, Max, and Enterprise customers through June 22. Three days of availability, then the plug gets pulled on what was supposed to be a goodwill launch. Existing sessions are now throwing errors, new sessions fall back to Opus 4.8, and developers using the API have been told to migrate.

For those unfamiliar with the model lineup: Fable 5 and Mythos 5 share the same underlying architecture, but Fable is the safety-filtered version aimed at general use, while Mythos is the unrestricted variant gated to vetted government cyber defenders and life sciences partners. Fable applies filters on sensitive cybersecurity, biology, and chemistry queries. Mythos does not.

As for why the government acted now, Anthropic's own account is revealing. The company says it believes the directive stems from a reported jailbreak of Fable 5. After reviewing the demonstration, Anthropic concluded it amounted to minor, already-known issues, the sort of thing that can be reproduced with other publicly available models without any special bypass technique.

In its public statement, Anthropic was fairly direct about its position: "To date, the government has only given us verbal evidence of a potential narrow, non-universal jailbreak, which essentially consists of asking the model to read a specific codebase and fix any software flaws." The company says it disagrees that a narrow potential jailbreak justifies pulling a commercial model deployed to hundreds of millions of people, and argues that if this standard were applied consistently across the industry, new frontier model deployments would effectively grind to a halt.

Anthropics also points out that the underlying capability is widely available elsewhere, including OpenAI's GPT-5.5, and is routinely used by security defenders.

The company says it is complying with the directive while simultaneously disputing its basis, and has promised further details within 24 hours. It frames this as a misunderstanding it hopes to resolve quickly.

Meanwhile, the UK's Minister for AI and Online Safety, Kanishka Narayan MP, weighed in to note the pause affects British customers too, using the moment to flag the government's £1.1bn AI chip investment as justification for pushing technological sovereignty. Whether that was helpful or just opportunistic positioning is left as an exercise for the reader.

All other Anthropic models, including Claude Opus 4.8, remain operational.