NSA spies are reportedly using Anthropic’s Mythos, despite Pentagon feud

The National Security Agency appears to be among the undisclosed recipients of Anthropic's Mythos model, according to a TechCrunch report. The NSA is reportedly using the model primarily for scanning environments to identify exploitable vulnerabilities. This development follows the Department of Defense's recent designation of Anthropic as a 'supply chain risk.'

Anthropic has not commented on this specific deployment. The Pentagon's risk designation, which typically signals concerns about procurement from certain vendors, does not appear to have prevented the NSA's use of the model. The report does not specify when the NSA began using Mythos or the exact terms of access.

The situation highlights potential inconsistencies in how different U.S. government agencies assess and adopt AI technologies from the same vendor. It also raises questions about the practical impact of supply chain risk designations on intelligence community procurement.