About the Paper
Modern military headquarters generate enormous volumes of operational data, yet the challenge increasingly lies not in collecting information but in converting it into trusted understanding at the speed of command. The Model Should Be the Smallest Part of the System: Hierarchy-Aware Axiomatic-Agentic Intelligence for Military Decision Support examines this challenge through the practical evolution of an AI-enabled military intelligence assistant. The paper argues that Artificial Intelligence should not establish truth, but reason over truth established through authoritative data and deterministic computation. It proposes a hierarchy-aware architecture in which security is enforced before data retrieval, specialist agents perform bounded analytical tasks, institutional knowledge grounds reasoning, and human authority governs learning and correction. The paper presents two complementary systems for direct operational queries and deeper intelligence assessment, offering a framework for secure, explainable and human-governed AI-enabled military decision support.
