AI-Driven Compliance and Configuration Intelligence at Scale: An Explainable, Human-Centered Framework for Enterprise Infrastructure

Abstract

Enterprise IT environments increasingly span heterogeneous on-premises, cloud, and hybrid platforms, where configuration changes occur continuously and at scale. Ensuring configuration and security compliance under these conditions remains a persistent challenge, particularly in regulated domains where transparency, auditability, and human accountability are mandatory. Traditional rule-based compliance mechanisms, while effective for baseline enforcement, struggle to provide timely detection of configuration drift, contextual risk assessment, and decision support. This paper proposes an AI-driven, explainable compliance and configuration intelligence framework designed to augment existing configuration management and governance tooling. Rather than replacing deterministic controls, the framework introduces data-driven risk inference, temporal drift analysis, and human-in-the-loop validation. We present a reference architecture derived from large-scale enterprise deployments and align it with established research in configuration management, explainable artificial intelligence (XAI), and human-AI interaction. The results demonstrate how explainable AI can enhance situational awareness, reduce compliance overhead, and improve governance without compromising control or trust.