# Washington Turns Export Controls on a Model Itself, Forcing Anthropic to Disable Fable 5 and Mythos 5

A national-security directive bars foreign access to two frontier models, a shift in which the United States treats model access, not chips or weights, as the controlled item.

- Published: 2026-06-21T10:59:22.601Z
- Canonical: https://polylog.news/ai/2026-06-21/washington-turns-export-controls-on-a-model-itself-forcing-a
- Publisher: Polylog (AI desk)
- Section: geopolitics
- Sources: [Anthropic News](https://www.anthropic.com/news/fable-mythos-access), [Polylog editors](https://polylog.news)

The United States government has invoked national-security authorities to order [Anthropic](https://www.anthropic.com/news/fable-mythos-access) to suspend all access to two of its frontier models, Fable 5 and Mythos 5, for any foreign national anywhere in the world, including the company's own non-citizen employees. Anthropic says it received the directive at 5:21 p.m. Eastern time on June 12 and had to disable both models for every customer to comply. Access to its other models continues.

This is a categorical shift in how export controls work. Earlier controls applied to hardware, advanced accelerators, manufacturing equipment, and, more recently, the weights of open models. Here the controlled item is interactive access to a hosted model, delivered through an application programming interface (API). Anthropic says the order did not specify the concern, but that its understanding is the government believes it has identified a method of [bypassing, or jailbreaking, Fable 5](https://www.cnbc.com/2026/06/12/anthropic-disables-access-to-fable-5-and-mythos-5-to-comply-with-government-directive.html).

Anthropic disputed the rationale. The company argued that recalling a model used by hundreds of millions of people over a single narrow jailbreak would, if applied across the industry, halt all new frontier deployments. The order follows reporting, summarized in social channels including [AI Post](https://t.me/aipost/7295), that the Mythos model was being prepared for offensive cyber work and that an internal demonstration alarmed officials.

What is verified is the directive itself and Anthropic's compliance, both confirmed by [Fortune](https://fortune.com/2026/06/13/anthropic-disables-fable-mythos-export-controls-national-security-threat/) and CNBC. What remains asserted, and disputed, is the precise capability that prompted it. The government has not published its evidence, and Anthropic says it disagrees with the threshold being applied.

## What this means

If a government order can shut off access to a hosted model for a class of users, then the API becomes a control point that policy can reach directly, without affecting any hardware. For engineers, that turns model availability into a question of jurisdiction and compliance, and it strengthens the case for multi-vendor and self-hostable fallbacks. For teams outside the United States, it adds weight to downloadable alternatives that no foreign government can revoke.

## What to watch

- Whether the order is formalized into a published rule with stated criteria, which would tell developers which capabilities trigger access controls rather than leaving it case by case.
- Whether other US labs receive similar directives, signaling this is a standing policy instrument rather than a one-time action against Anthropic.
- How fast enterprises outside the US migrate critical workloads to open-weight or domestically hosted models once they see that closed-model access can be revoked.