The U.S. National Security Agency (NSA) is evaluating Anthropic’s advanced AI model, Mythos, to identify cybersecurity flaws in commercial software. Testing specifically targets products from Microsoft. NSA officials report the model demonstrates superior speed and efficiency compared to traditional discovery methods.
Anthropic maintains Mythos as a private tool to prevent potential misuse by malicious actors. The AI reportedly identifies zero-day vulnerabilities within major operating systems.
This evaluation follows the Pentagon’s previous designation of Anthropic as a potential supply chain risk. The testing signals a complex shift in the relationship between the U.S. government and the AI developer.