Washington’s Midnight Directive: Why Anthropic’s Fable 5 and Mythos 5 Were Pulled Offline

In a sudden regulatory move, the US government has forced Anthropic to suspend its most advanced AI models, triggering a fierce debate over national security and transparency.

The clock struck 5:21 p.m. ET on a Friday when the regulatory hammer dropped. In a move that has sent ripples through the global tech ecosystem, the US government issued an urgent export control directive to Anthropic, effectively mandating an immediate shutdown of its cutting-edge Fable 5 and Mythos 5 models. The order is sweeping: it prohibits access to these specific systems for any foreign national, regardless of their location, forcing the company to abruptly pull the plug for its global user base.

The Company’s Pushback

Anthropic did not take the intervention lying down. In a blistering 700-plus word statement, the company confirmed it was complying with the directive while simultaneously challenging the government’s logic. According to the firm, the authorities flagged a potential "jailbreak" method—a way to bypass safety guardrails—but provided little in the way of concrete detail. Anthropic contends that the vulnerability identified was narrow, minor, and already known, arguing that it hardly warrants a total recall of a product deployed to millions.

For a sector that has operated with significant latitude, this incident marks a sharp departure from traditional oversight. Anthropic’s leadership has publicly urged the government to adopt a more formal, statutory process. They argue that if the state intends to block commercial deployments, it must do so through a framework that is transparent, technically grounded, and fair—rather than through abrupt, high-stakes mandates that catch companies off guard.

Why it matters

The fallout from this directive is far-reaching. Beyond the immediate disruption for users, it signals an era where national security concerns are increasingly being weaponised to control the distribution of powerful digital infrastructure. By invoking export controls to restrict AI access, the US is essentially treating advanced language models like sensitive dual-use technology, akin to semiconductors or advanced encryption.

This sets a precarious precedent. When governments use opaque directives to stifle commercial software, it creates a "chilling effect" on innovation. Startups and tech giants alike are now forced to navigate a geopolitical minefield where a single security report could lead to the overnight loss of a core product. For the global market, the message is clear: the era of unchecked AI deployment is over, and the regulatory state is moving in with heavy-handed speed.

The Bigger Picture

What makes this case particularly thorny is the lack of clarity. If the government holds evidence that these models pose a genuine existential threat, the public—and the developers—deserve a clear explanation. By failing to provide specific details, the state risks appearing arbitrary. As this situation develops, the industry will be watching closely to see if this is an isolated incident or the first move in a broader, more aggressive US strategy to restrict AI access on the world stage.