NSA Integrates Anthropic's Mythos Preview Despite Pentagon Risk Warnings

The National Security Agency (NSA) has begun utilizing 'Mythos Preview,' the latest advanced artificial intelligence model developed by Anthropic. This deployment comes amid reported internal friction within the U.S. defense and intelligence community regarding the security of the provider. According to reports, senior officials at the Pentagon continue to categorize Anthropic as a potential supply-chain risk. This designation suggests a fundamental disagreement between the NSA and other defense leadership on the acceptable risk profile for integrating third-party large language models (LLMs) into sensitive government operations. The NSA's decision to move forward with Mythos Preview underscores the agency's priority to leverage cutting-edge AI capabilities for intelligence gathering and analysis, even when facing regulatory or security warnings from other branches of the defense establishment. While Anthropic remains a private entity, this adoption signals strong institutional demand for high-tier AI models. The situation highlights the ongoing tension between the urgent need for AI modernization and the stringent security protocols required for national security infrastructure.