Policy as an Acceleration System: The Doctrine of Speed, Institutional Engineering, and Auditable Execution Mechanisms in the US Department of War’s ‘War Mobilisation-Level Artificial Intelligence Strategy’

This paper proposes and validates a computable interpretative framework for ‘policy as an acceleration system’, using as its sole material the policy directive text issued by the US War Department on 9 January 2026: the Memorandum from Senior Leadership at the Pentagon: Artificial Intelligence Strategy for the Department of War. The author treats this text as an institutional acceleration blueprint at the level of war mobilisation: it constructs a self-reinforcing deployment acceleration loop through compressible time constraints, a portfolio of benchmark projects, foundational enablement centred on data and procurement, and a governance mechanism paced by monthly review and demonstration milestones. Methodologically, this paper operates a military-grade, reproducible, and auditable algorithmic pipeline: Text → Event Table → Institutional Variable Dictionary → Mechanism Chain and Causal Candidate Graph → Multi-layer Network → Velocity Doctrine Index (VDI) → Robustness and Audit Table. Results demonstrate that the institutional design described herein transcends mere ‘innovation encouragement,’ instead institutionalising velocity as a measurable, accountable, and iterable organisational capability: PSP operates under standards of single accountability, radical timelines, measurable outcomes, and rapid iteration. By exposing data catalogues/interfaces and mandating 30-day updates, it significantly reduces data friction, thereby enhancing trial-and-error density and diffusion velocity. This paper contributes: a verifiable policy acceleration system modelling language; a reusable event-variable-mechanism framework; and a comparable VDI index with network bottleneck diagnostics. These provide verifiable computational evidence for national-level AI governance and organisational innovation research.