The Trump administration is rolling back export controls on Anthropic’s most advanced AI models, Fable 5 and Mythos 5. This decision follows a recent agreement between Anthropic and the Commerce Department, as communicated in a letter from Commerce Secretary Howard Lutnick to Anthropic co-founder Tom Brown.
Previously, these AI models were only approved for release to select companies and governmental bodies due to security concerns. In the new directive, Lutnick stated that a license is no longer required for the models’ export or in-country transfer.
These changes come after Anthropic has been collaborating with the Commerce Department to bolster safeguards against potential misuse, particularly concerning cybersecurity features. Anthropic has committed to proactively identifying and addressing security risks linked to the models and will work closely with the U.S. government on relevant protocols.
Initially, Anthropic was skeptical about the administration’s concerns, arguing that it was impossible to eliminate all potential "jailbreak" attempts that could exploit functionalities of the Mythos model. However, the company has shifted its approach to regain access for Fable. This included changing leadership dynamics during discussions, with Brown taking over more prominent roles from CEO Dario Amodei, who officials found less agreeable.
Despite these changes, there is still skepticism about whether Anthropic can effectively curb the jailbreaks that threaten the integrity of their models.
