Next-Generation Zero-Trust Identity Orchestration for Unified Human–Machine Access in Critical Infrastructure and Healthcare Networks

Abstract

Background: Rapid convergence of cyber-physical systems (CPS), Internet of Medical Things (IoMT), and distributed cloud/edge services has created environments where human and machine identities coexist and interact continuously. Traditional perimeter-based security assumptions are no longer tenable; zero-trust architecture (ZTA) shifts the model to identity- and policy-centric access decisions (Rose et al., 2020; Kindervag, 2010). Despite substantial literature on ZTA components and industrial adaptations (Stafford, 2020; Syed et al., 2022; Feng & Hu, 2023), there is limited work that integrates intent-awareness — the capacity to interpret, represent, and enforce access based on the actor’s operational intent — across heterogeneous identity types (human users, service accounts, IoT devices, ML agents) within safety-critical domains such as healthcare.

 Objective: This paper proposes a comprehensive, publication-ready design and evaluative narrative for an Intent-Aware Zero-Trust Identity Architecture (IA-ZTIA) that unifies human and machine access control. The architecture is grounded in contemporary ZTA guidance (Rose et al., 2020; Stafford, 2020), research surveys (Syed et al., 2022; Yan & Wang, 2020), and domain-specific constraints from healthcare and CPS literature (Loh et al., 2022; Feng & Hu, 2023; Lakhan et al., 2022).

 Methods: We describe a systems-level methodology articulating identity modeling, intent representation, continuous trust evaluation, policy orchestration, telemetry and explainability, and privacy-preserving learning for intent inference. Methods are textually specified to be implementable without formulae: design choices, data flows, trust scoring semantics, and governance mechanisms are described in depth. We synthesize evidence from existing ZTA deployments (Gilman, 2016; Osborn et al., 2016) and healthcare-AI and federated learning work (Islam et al., 2023; Lakhan et al., 2022) to justify design decisions.

 Results: The descriptive analysis identifies how IA-ZTIA improves risk differentiation, reduces lateral movement opportunities, and enables safe machine-to-machine delegation patterns while preserving patient privacy and auditability. We document qualitative outcomes: finer-grained authorization, improved incident forensics, reduced blast radius for compromised machine identities, and compatibility with regulatory privacy requirements. Trade-offs around latency, complexity, and model overfitting in intent classifiers are analyzed.

 Conclusions: Intent-awareness augments ZTA by aligning access decisions with operational context and purpose, particularly valuable in CPS and healthcare where actions have physical consequences. We provide a roadmap for staged adoption, governance recommendations, and areas requiring future empirical validation, including real-world deployment studies and longitudinal safety evaluations. This article synthesizes interdisciplinary knowledge into a coherent architecture and critical discussion for researchers, engineers, and policy-makers seeking to unify human and machine identity under zero-trust principles. (Max. 400 words) (Rose et al., 2020; Stafford, 2020; Syed et al., 2022).