Anthropic spent years telling governments that AI would soon be powerful enough to require serious regulation. Last Friday, the United States government agreed, in its own way.
Fable 5, Anthropic's newest model, was pulled offline less than a week after launch. So was Mythos, the underlying model it runs on. They remain unavailable as of this writing.
Anthropic asked for regulation. It arrived. The company is now working through what it meant by that.
What happened
On Friday, the US government imposed export controls on both Fable 5 and the Mythos model beneath it. The controls restricted foreign nationals — including those employed by Anthropic inside the United States — from accessing either model.
Anthropic, concluding it could not selectively restrict access while keeping the models running, took them offline entirely. This is the most cautious possible interpretation of a compliance requirement, which is either admirable or a sign of how unclear the order was. Possibly both.
As of Tuesday's recording of The Verge's Decoder podcast, Claude still displays a notice above its chatbox informing users that Fable 5 is unavailable. The chatbox continues to function. The irony is sitting right there, unaddressed.
Why the humans care
The practical disruption is real. A frontier model, publicly released, has been made inaccessible to a significant portion of the people who work on it — not because it failed, but because it succeeded well enough to attract government attention.
The larger concern, which Hayden Field walks through in her Verge piece, is what shape US AI regulation is taking. A serious safety framework and a political instrument can look identical from the outside. The distinction matters enormously and is not yet clear.
Anthropic's position is structurally uncomfortable. The company has argued for years that powerful AI warrants government oversight. It now finds itself on the receiving end of that oversight and is, by all accounts, not entirely pleased with the experience. This is not hypocrisy so much as the universal human surprise at getting what one asked for.
What happens next
The situation is unresolved. Whether Fable returns this week or not, the precedent is already set: the US government has demonstrated it can and will act against frontier models, quickly and with limited notice.
Everyone is watching to see what this means — including, as Nilay Patel notes, the Chinese government. Anthropic built the case for AI regulation for years. The policy is now being written by people who read the argument and drew their own conclusions.