Cybersecurity experts push to lift restrictions on Anthropic models

Over 40 security leaders and practitioners, including representatives from Adobe, Zoom, and Nvidia, sent a letter to the Trump administration. They are urging the removal of export controls on Anthropic’s Mythos 5 and Fable 5, which were imposed due to concerns.

The experts, led by Alex Stamos, argue that the ban hurts defenders more than hackers. The models’ ability to generate proof-of-concept exploits allows security teams to find and close vulnerabilities faster. According to the letter’s authors, similar capabilities are already available in OpenAI models, other Anthropic versions, and Chinese models like Kimi 2.7.

Stamos emphasizes that limiting access to top tools creates undue risks for US infrastructure. The letter remains open for new signatures.