Anthropic AI Engineers Train NSA on Mythos for Cyber Ops

Anthropic has reportedly sent Mythos AI engineers to the National Security Agency, or NSA. An anonymous source told the Financial Times that these engineers are helping the NSA adopt and train with the powerful Mythos AI. What's interesting is this suggests the government has waived its previous "supply chain risk" designation for Mythos AI in some intelligence operations. The Department of War banned Anthropic in March after the company stated it wouldn't support the AI for autonomous weapons or domestic surveillance. About "half a dozen" Mythos AI engineers have been dispatched to assist the NSA with "certain applications." This isn't the NSA's first interaction with Mythos AI; reports in April indicated the agency was already quietly using it despite the ban. The NSA is likely exploring Mythos AI's vulnerability scanning capabilities. Anthropic's Project Glasswing has shown impressive results in quickly identifying hidden vulnerabilities. This development could significantly impact how organizations manage cybersecurity.