DoD AI Requirements & Strategy
AI Acceleration Strategy (January 2026)
On January 9, 2026, the Department of Defense released its AI Acceleration Strategy, with the stated goal of making the U.S. military “AI-first.” The Secretary of Defense’s memo lays out four aims: (1) incentivize internal DoD experimentation with AI models; (2) identify and eliminate bureaucratic obstacles to model integration; (3) focus investment on asymmetric advantages — AI computing, model innovation, entrepreneurial dynamism, capital markets, operational data; (4) launch pace-setting projects (PSPs) — tangible, outcome-oriented vehicles for rapidly building AI infrastructure, data, models, policies, and talent.
The seven pace-setting projects
Warfighting: Swarm Forge (competitive human-machine teaming), Agent Network (AI agents for battle management), Ender’s Foundry (AI-enabled simulation).
Intelligence: Open Arsenal (technical-intelligence-to-capability pipeline), Project Grant (dynamic deterrence with interpretable results).
Enterprise: GenAI.mil (democratizing AI experimentation at all classification levels).
GenAI.mil
The Pentagon’s generative AI platform, launched December 2025 and built in 74 days. Nearly 2 million users across the department — service members, civilians, contractors. Started with Google’s Gemini; now includes OpenAI’s ChatGPT and xAI’s Grok. Certified for CUI at Impact Level 5. DoD is driving adoption through training and no-cost command-level onboarding.
The organization — CDAO
The Chief Digital and Artificial Intelligence Office runs the portfolio — and was folded under the Under Secretary of Defense for Research and Engineering, giving it direct budgetary authority and a shorter line to the Secretary’s AI roadmap.
Related requirements
EO 14409 (June 2026): expand AI-enabled cyber defenses; build a classified process for assessing frontier AI models’ cyber capabilities.
Multi-vendor posture: as of May 2026, seven providers cleared for classified networks — Google, Microsoft, AWS, SpaceX, Oracle, OpenAI, Reflection. Policy explicitly warns against single-vendor dependency.