Understanding Before Interpretation: A Structural Framework for Decision Support and AI Governance

This paper proposes a structural interpretation framework addressing a persistent limitation in explainable artificial intelligence (XAI) and decision-support systems: the disconnect between explanation, user understanding, and actionable decision-making. It introduces the Understanding Constraint (UC) and the Interpretation Alignment Principle (IAP), and proposes a Three-Layer Interpretation Model (Understanding–Interpretation–Action) as a pre-execution structural condition for governance. The work contributes to interpretability research, human-in-the-loop decision systems, and AI governance by reframing interpretation as a governed structural layer rather than a descriptive process.