It has been nearly a week since the Trump administration ordered Anthropic to pull its most advanced models offline. Claude Mythos 5 and Fable 5 remain suspended. Anthropic leaders have flown to Washington, met with White House officials, and negotiated for days. And yet no one can say exactly what the company did wrong (Wired, June 18).
The standoff has become a case study in what happens when the government tries to regulate frontier AI without a clear legal framework. The export control directive, issued under national security authorities, bars “any foreign national” from accessing the models, including foreign national Anthropic employees. The company complied immediately, but neither side can articulate the specific rule Anthropic violated.
### The Core Tension
The White House believes it has identified a method of bypassing Fable 5’s safety guardrails, according to an earlier blog post from Anthropic. The government’s concern centers on the model’s potential for cyberattack capabilities, which is why the export controls target foreign national access.
But the administration has not published a clear standard for what constitutes a violation or what remedy would satisfy its concerns. The result is regulatory limbo: the models cannot be distributed, the company cannot fix a problem it does not fully understand, and the government cannot point to a specific rule that was broken (Wired, June 18).
This marks a shift from earlier reporting on the saga. The initial narrative centered on whether Mythos 5 could be used for offensive cyber operations. By day six, the question has become whether the US government has any coherent AI policy at all.
### How We Got Here
The crisis began June 12, when Anthropic announced it had received an export control directive from the US government citing national security authorities. The order forced the company to disable Fable 5 and Mythos 5 for all customers to ensure compliance, including blocking access for foreign national employees of Anthropic itself (Anthropic/X, June 12).
Days of high-level talks between Anthropic leadership and White House officials failed to resolve the impasse. Reports emerged that a China-linked group may have accessed Mythos before the shutdown, though the details remain unclear (CNBC, June 15). Cybersecurity experts wrote to the White House urging restraint (AP News, June 16).
The trigger appears to be rooted in the export control directive itself, which the US government issued after identifying a method of bypassing Fable 5’s safety guardrails, according to Anthropic’s initial statement.
### The Bigger Problem
The Anthropic case is not an isolated incident. It exposes a structural gap in US AI governance. The Trump administration killed the previous executive order on AI safety early in its term, leaving no statutory framework for export controls on AI models. The government is now governing by directive, improvising restrictions on a case-by-case basis.
The result creates problems on both sides. Companies cannot plan for compliance when the rules are invented after the fact. The government cannot credibly claim to have a coherent AI strategy when it cannot articulate what law was broken. And the AI models remain offline, a practical loss for enterprise customers who had integrated them into their workflows.
For Anthropic, the timing is particularly painful. The company confidentially filed for what could be the largest IPO in history on June 1, 2026 (Anthropic; Wired), just weeks before the export control crisis. Whether regulators will clear the models before the IPO roadshow begins is anyone’s guess.
This is a follow-up in the ongoing Anthropic export control saga.
Sources: Wired (June 18); CNBC (June 15); Anthropic/X (June 12)