Cybersecurity Experts Urge U.S. to Lift Export Ban on Anthropic’s AI Models

A coalition of 76 cybersecurity experts, including notable industry figures like Alex Stamos and Katie Moussouris, has published an open letter urging the U.S. government to reverse the export ban on Anthropic’s Fable and Mythos AI models. The government imposed this restriction last Friday, citing unspecified national security concerns, prompting Anthropic to suspend global access to these models.

• The experts argue that the export control hinders cybersecurity defenders by removing powerful tools essential for identifying software vulnerabilities.
• Anthropic initially limited access to its AI models to prevent misuse by malicious entities, but these restrictions are now seen as obstacles by the security community.
• Katie Moussouris has critically reviewed a private Amazon paper that allegedly influenced the export ban. She contends that the paper’s findings were not indicative of a real vulnerability in the models.
• The letter advocates for a more transparent and democratic regulatory process, informed by scientific research and minimally invasive measures for public safety.
• The signatories note that similar capabilities can be achieved using other AI models, including those from OpenAI and Chinese developers.