About

Anthropic sends staff to Washington to renegotiate AI export controls

Published
Score
10

Why it matters

Anthropic has sent senior technical staff to Washington to negotiate with White House and Commerce Department officials over export restrictions that forced the company to take its most advanced AI models offline. The dispute centers on Anthropic's Mythos and Fable models, which the Trump administration ordered restricted from access by foreign nationals—including those physically present in the U.S.—citing national security and cybersecurity concerns. Commerce Secretary Howard Lutnick and Under Secretary Jeffrey Kessler are leading the government side, while Anthropic CEO Dario Amodei is directing the company's response. The administration's restrictions came after claims that one of the models had been jailbroken, with the National Security Agency reviewing findings that informed the government's position.

The timeline and scope of the restrictions remain in flux. Anthropic disabled both models for all customers while attempting to comply with the directive, but the precise technical requirements the government is imposing and the duration of the restrictions are not yet clear. The company says it has been cooperating with the government's orders, though administration officials have publicly disputed the characterization of Anthropic's compliance efforts.

This clash signals how the Trump administration intends to police frontier AI systems domestically. For attorneys advising AI companies, the episode demonstrates that export controls can be weaponized rapidly and with broad effect—forcing companies offline rather than merely restricting foreign access. The outcome of Anthropic's negotiations will likely set a precedent for how the government handles similar disputes with other AI developers, making the resolution worth tracking closely.

Sources

mail Subscribe to Artificial Intelligence email updates

Primary sources. No fluff. Straight to your inbox.

Also on LawSnap