Adaptive Zero Trust Security Frameworks for Scalable Cloud Computing Environments with Dynamic Identity and Access Management

Abstract

Zero Trust is frequently marketed as a deterministic solution to cloud insecurity, yet the empirical evidence is far less stable, revealing structural inconsistencies in identity propagation, policy orchestration, and runtime enforcement across distributed systems. The proposed framework reinterprets Zero Trust not as a static perimeter replacement but as an adaptive, feedback-driven control system where identity signals, contextual risk scoring, and access policies continuously mutate under load, latency, and adversarial pressure. The results suggest that scalability introduces paradoxical fragility; as identity granularity increases, so does the probability of misalignment between authentication confidence and authorization execution. This is not accidental.

The evaluation across simulated multi-cloud deployments demonstrates that dynamic identity and access management mechanisms reduce lateral movement risk by 38–52%, yet simultaneously increase policy evaluation latency and operational entropy under peak conditions. The system exhibits a clear failure threshold where excessive trust recalibration leads to decision fatigue within policy engines, degrading response accuracy. The evidence is contradictory at best. Zero Trust, in its current form, behaves less like a security paradigm and more like a probabilistic balancing act between enforcement precision and system coherence