AI / ML / Autonomous Systems — 2026-03-17

Pentagon Orders Removal of Anthropic AI from Key Military Systems After Supply Chain Risk Designation

An internal Pentagon memo dated March 6 designated Anthropic's AI as an 'unacceptable supply chain risk' after negotiations collapsed over two categorical restrictions: Anthropic refused to power fully autonomous lethal weapons or domestic mass surveillance systems. Military commanders must now remove Anthropic AI from nuclear weapons, ballistic missile defense, and cyber warfare systems. OpenAI secured the replacement Pentagon contract amid criticism for not releasing full terms.

Analysis
Prior INTSUMs tracked the Anthropic-Pentagon rupture from initial blacklisting (Mar 6) through supply chain risk designation. The six-month removal timeline creates a vulnerability window: military systems must migrate to OpenAI alternatives while maintaining operational continuity during an active war.