The US government just did something it has never done before in the commercial AI sector: it told one of the world’s most valuable companies to cut off foreign users from its best products, effective immediately.
On June 12, Anthropic disabled access to its Fable 5 and Mythos 5 models for all foreign nationals, following a directive from the US Department of Commerce. The trigger was a high-risk jailbreak vulnerability that CEO Dario Amodei flagged in a blog post, which quickly escalated into a national security conversation involving tech leaders including Amazon CEO Andy Jassy.
What happened, and why it matters beyond AI
The Commerce Department’s directive represents one of the most significant government interventions in the commercial AI sector to date. Think of it as the AI equivalent of export controls on advanced semiconductors, except this time the product being restricted isn’t hardware. It’s intelligence itself.
Anthropic, valued at somewhere between $965 billion and approximately $1 trillion, isn’t some scrappy startup that can be quietly sidelined. It’s a cornerstone of the global AI infrastructure.
