Anthropic's most powerful AI models remain offline two weeks after the Trump administration issued an unprecedented export control order, and there is no resolution in sight.

The standoff began on June 12 when the administration demanded Anthropic suspend access to its Mythos 5 and Fable 5 models by "any foreign national" — including non-US citizens employed by Anthropic itself — citing security concerns. The order came after the National Security Agency concluded there were ways to disable Fable 5's guardrails, despite the Commerce Department having tested the model before release and raising no objections.

Anthropic co-founder Tom Brown has now replaced CEO Dario Amodei in negotiations, but after 14 days of intensive talks, the two sides remain deadlocked. The administration views the situation as Anthropic's problem to fix, according to three people familiar with the discussions, while independent cybersecurity experts increasingly argue that what the White House is demanding — blocking all possible jailbreaks — may be technically impossible.

The prolonged shutdown threatens the US AI industry. Anthropic needs Mythos-class revenue to fund its $15 billion-per-year SpaceX data center deal and maintain momentum ahead of what could be the largest IPO in history. OpenAI confirmed it will delay its GPT-5.6 release after a similar administration request, and countries abroad have begun calling for non-American AI alternatives. As cybersecurity expert Alex Stamos put it: "One of America's champions is being kneecapped by the US government while we're in a race with the Chinese."

Sources: The Verge, Wired