Trump Administration Lifts Export Controls on Anthropic’s AI Models
The Trump administration is set to lift export restrictions on Anthropic's most advanced AI models after a deal with the Commerce Department, easing earlier controls that had suspended access for foreign nationals.
The White House is expected to announce Tuesday night that it will lift export controls on Anthropic's two most powerful AI models: Fable 5 and the more advanced Mythos 5, according to a person familiar with the matter. Mythos 5 had previously been approved only for select companies and government agencies.
The decision will be communicated in a letter from Commerce Secretary Howard Lutnick to Anthropic cofounder Tom Brown. Lutnick, alongside national cyber director Sean Cairncross, has been leading the administration's efforts to resolve the dispute with Anthropic.
The developments come as Anthropic has been collaborating with the Commerce Department and the White House to strengthen safeguards against users bypassing Fable's safety restrictions, particularly in cybersecurity. Initially, Anthropic argued that the administration's jailbreak concerns were overblown, stating it was impossible to ensure zero jailbreaks that could unlock the restricted capabilities of Mythos.
In recent weeks, Anthropic changed its approach by trying to restore Fable's availability, which also meant altering its communication style with the administration. Reports indicate that Anthropic CEO Dario Amodei was replaced in meetings by cofounder Tom Brown, who officials found more personally agreeable. Anthropic assured the administration it would reduce jailbreak incidents by building more robust safeguards, effectively telling officials what they wanted to hear rather than debating whether jailbreaks could be entirely stopped.
The move reverses earlier orders that suspended access for foreign nationals to Anthropic's advanced models. Now, after reaching a deal, the restrictions are being lifted.
