The US government just did something it has never done before: it used export control powers to shut down global access to a specific company’s AI models. Commerce Secretary Howard Lutnick sent a directive to Anthropic CEO Dario Amodei on June 12, 2026, ordering the immediate suspension of exports and access to the company’s Fable 5 and Mythos 5 models for all foreign nationals.

By June 13, Anthropic had complied, disabling global access to both models. The speed of the response, less than 24 hours, tells you everything about how seriously the company is taking this.

What happened and why it matters

Here’s the thing. Fable 5 had been publicly released just three days earlier, on June 9. Three days from launch to global shutdown is not a normal product lifecycle.

The directive marks the first time the 2018 Export Control Reform Act has been applied specifically to AI technology. That law was originally designed to give the Commerce Department authority over emerging and foundational technologies that could threaten national security. Until now, its most visible applications have targeted semiconductor exports to China.

Lutnick’s letter identified a specific risk: that Anthropic’s advanced models could be diverted to military intelligence operations in nations like China and Russia.

Anthropic, for its part, didn’t just flip the off switch and stay quiet. The company raised its own concerns about the practical challenges of enforcing nationality-based restrictions on cloud-based services. The company also flagged the persistent problem of “jailbreaks,” techniques that allow users to bypass built-in safety guardrails.

The backstory runs deeper than one letter

This confrontation didn’t materialize overnight. The tensions between Anthropic and the US government trace back to February 2026, when the company refused to support military applications of its models. Specifically, Anthropic drew a line against domestic surveillance and autonomous weapons use cases.

That refusal led to a notable consequence: the Pentagon designated Anthropic as a “supply chain risk.”

What this means for investors

For AI companies, the message is clear: if your models are powerful enough to matter, they’re powerful enough to regulate. Companies that have built their growth strategies around global access, selling cloud-based AI services to customers in every jurisdiction, now face a new category of risk. A single government letter can shut down international operations overnight.

Tech industry leaders are already responding. Executives from companies including Nvidia and Adobe have voiced concerns about what these restrictions mean for innovation and competitiveness.

For crypto markets specifically, this episode doesn’t have a direct connection. No tokens or digital assets are tied to Anthropic’s situation.

Anthropic’s compliance, while swift, also raises questions about centralization risk in AI. A single directive from a single government official was enough to disable a product globally.