The US government has moved to restrict access to Fable 5, the artificial intelligence model that Anthropic had only recently released to the public.
According to the Associated Press, Defense Secretary Pete Hegseth issued an ultimatum to Anthropic this week, demanding the company open up its artificial intelligence. The AP frames the standoff around the Defense Production Act, the federal law that lets the government compel private companies to prioritize national-security needs.
The practical effect was swift. PCMag reports that Fable 5 — which it describes as Anthropic's most powerful model ever — disappeared just days after its launch. One of the outlet's writers said they managed to test the model's capabilities during the brief window it remained available, before the government shut it down.
Taken together, the two accounts describe a rare collision between a leading AI developer and the Pentagon: a brand-new commercial product pulled from general availability almost as soon as it appeared, under pressure tied to a wartime-era industrial law.
The sources here do not spell out exactly what the Pentagon wants Anthropic to hand over, how the company has responded, or how long any restriction will last. What is clear from the reporting is that the most capable model Anthropic had shipped is no longer freely accessible, and that the demand came directly from the top of the Defense Department.
Why it matters: When the government can invoke a national-security law to limit who uses the most advanced commercial AI, it signals that frontier models are now treated as strategic assets — reshaping the relationship between Silicon Valley's AI labs and the state.