The Trump administration is rolling back export controls on Anthropic's advanced AI models, Mythos and Fable, following a resolution with the Commerce Department.
Commerce Secretary Howard Lutnick informed Anthropic co-founder Tom Brown in a letter that restrictions on both the Fable 5 and the more potent Mythos 5 models are now lifted. Previously, these models were only accessible to select companies and government bodies. Lutnick stated, "A license is no longer required for the export, reexport, or in-country transfer, including deemed export or deemed reexport, of the Mythos or Fable models."
This move comes as Anthropic has been collaborating with the Commerce Department and the White House to bolster safety measures. The goal is to prevent users from bypassing Fable's security protocols to access restricted functionalities, particularly those related to cybersecurity.
According to Lutnick's letter, Anthropic has committed to proactively identifying and mitigating security risks tied to the models. The company will also work closely with the U.S. government on protocols, standards, and future releases for Mythos, Fable, and subsequent AI developments.
Initially, Anthropic argued that the administration's security concerns were exaggerated, suggesting it was impossible to completely eliminate "jailbreaks" that could unlock Mythos's restricted capabilities. However, the company has recently shifted its strategy, focusing on cooperation and enhanced safeguards to get Fable back online.
Sources indicate that Anthropic assured the administration it would prioritize reducing jailbreaks by implementing more robust safety features, effectively aligning its communication with the administration's stated priorities rather than debating the feasibility of stopping all jailbreaks.
