- Published on
The Trump Administration Is Lifting Its Export Controls on Anthropic’s Mythos and Fable AI Models
- Authors
- Name
- DeepHorus
The Trump administration is lifting export controls on Anthropic's two most advanced AI models, Mythos 5 and Fable 5, following a deal between the company and the Commerce Department. Commerce Secretary Howard Lutnick communicated the decision in a letter to Anthropic cofounder Tom Brown, obtained by WIRED. The move ends restrictions that had previously limited the models' release to select companies and government agencies only. This marks a significant shift in the administration's approach to regulating cutting-edge AI technologies.
Anthropic had faced U.S. government export controls on its powerful AI models due to national security concerns, particularly over potential misuse in cybersecurity. The company originally argued that the administration's worries were overblown, stating it was impossible to guarantee zero jailbreaks—techniques that bypass safety restrictions. However, in recent weeks, Anthropic changed its strategy to secure the lifting of restrictions, altering its communication style with officials.
Under the agreement, Anthropic committed to proactively detect and address security risks associated with the models, as well as work diligently with the U.S. government on protocols and standards for Mythos, Fable, and future models. Lutnick wrote that a license is no longer required for export, reexport, or in-country transfer of these models. The administration has been leading efforts to resolve the dispute through Lutnick and national cyber director Sean Cairncross.
The implications of this decision are substantial. By lifting export controls, the U.S. government enables broader global access to AI models that could accelerate innovation in various sectors but also raise security concerns. Anthropic's agreement to strengthen safeguards aims to mitigate risks, but critics may worry about potential misuse by hostile actors. The move reflects a compromise between fostering technological leadership and addressing security vulnerabilities.
Reactions from industry and policy circles are mixed. Some praise the administration for enabling U.S. AI companies to compete globally, while others question the effectiveness of voluntary safeguards. Notably, Anthropic replaced CEO Dario Amodei with cofounder Tom Brown in meetings with officials, who reportedly preferred Brown's personal style. This personnel change underscores the company's shift toward collaboration with the government.
Looking ahead, Anthropic will focus on implementing robust safeguards and reducing jailbreaks as part of its agreement. The company is expected to collaborate with the Commerce Department and the White House on future releases. This case sets a precedent for how the U.S. government might balance AI export controls with industry cooperation, potentially influencing policies for other advanced AI models.
