This paper presents an AI-driven, risk-adaptive Conditional Access framework designed to enhance identity governance in modern enterprise environments. Built on Microsoft Entra ID, the proposed model integrates identity, device posture, behavioral analytics, and threat intelligence signals into a unified decision-making system. Unlike conventional static policy enforcement, this approach enables continuous, context-aware access evaluation aligned with Zero Trust principles. The contribution of this work lies in the structured orchestration of multi-source signals into an adaptive access control model that improves both security resilience and operational efficiency.
Dilawar Shaikh (Sun,) studied this question.