Cybersecurity experts urge White House to lift export controls on Anthropic AI models

A coalition of 76 cybersecurity specialists, including former Facebook chief of security Alex Stamos and Bugcrowd founder Casey Ellis, sent an open letter to the White House requesting that export controls on Anthropic's Fable and Mythos AI models be rescinded. The signatories argue that the restrictions prevent defenders from using the models to identify and remediate software vulnerabilities, thereby reducing overall security.

The U.S. government ordered Anthropic to limit the export of the two models on national-security grounds, prompting the company to suspend worldwide access. Anthropic later released Fable, a version of Mythos with guardrails intended to block use in biology, chemistry and cybersecurity, but many experts say the safeguards also block legitimate security-related queries.

"To pull the best capabilities away from defenders without a good reason when our adversaries are rapidly advancing is dangerous," the letter states.

Katie Moussouris, founder of Luta Security and one of the letter’s authors, said a recent Amazon research paper cited as the basis for the export order did not demonstrate a true jailbreak. She explained that the paper simply asked Fable to fix code containing known and deliberately inserted vulnerabilities after the model initially refused to review the code. Moussouris contended that such behavior does not constitute a bypass and that defenders need AI that can locate, fix, and test security bugs.

The open letter also called for any regulations to be developed through a transparent, democratic process grounded in scientific research and applied only to the extent necessary to protect the public.