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

This dataset constructs and validates the computable interpretability framework ‘Policy as an Acceleration System’ using solely the policy directive text Artificial Intelligence Strategy for the Department of War, issued by the US Department of War on 9 January 2026. The dataset provides a reproducible, auditable algorithmic pipeline output comprising: page/sentence/ clause-level text slice corpus (with page-numbered evidence anchors), an institutional event table, an institutional variable dictionary, mechanism edge lists/chains, a multiplex network of nodes and edges, alongside a comparable **Velocity Doctrine Index (VDI)** with sensitivity analysis and audit matrix. This dataset enables researchers to quantitatively identify how policy texts form self-reinforcing deployment acceleration loops through time constraints, benchmark project portfolios, data and procurement enablement, and monthly review and demonstration rhythms. It supports cross-policy text comparability assessments and network bottleneck diagnostics, serving computational evidence research for national-level AI governance, institutional engineering, and organisational innovation.