RISK-ADAPTIVE ZERO TRUST ACCESS CONTROL UNDER CONTEXTUAL UNCERTAINTY: A SAFETY-CONSTRAINED FRAMEWORK FOR RELIABLE AUTHORIZATION

Abstract

While Zero Trust Architecture boosts network security by enforcing explicit authorization, the ability of adaptive access-control decisions to leverage missing, delayed or conflicting contextual information is hampered. This research introduces a Zero Trust framework that combines uncertainty-aware risk assessment and safety-constrained policy enforcement. A Design Science Research methodology will be used to guide the development of the framework, prototype development with Python and Open Policy Agent and controlled quantitative evaluation. The proposed framework will be compared with the static access control and conventional adaptive Zero Trust in various cases of contextual uncertainty. Evaluation metrics covers following areas: first unauthorized access accepted, false denial rate, classification rate, authorization latency, policy oscillation, and policy-invariant violations. Confidence intervals, hypothesis tests, effect sizes and sensitivity analysis will be used for statistical analyses. The study is geared toward answering the question of whether explicitly modeling contextual uncertainty with strict security constraints can enhance the security of authorizations while maintaining reasonable operational overhead.