AI-DRIVEN ZERO TRUST SECURITY FRAMEWORK FOR DETECTING AND PREVENTING CYBER THREATS IN MULTI-CLOUD ENVIRONMENTS

Abstract

Multi-cloud computing disperses applications, data, identities, workloads, and services throughout multiple cloud providers/administrative domains. While this architecture can be flexible, scalable, and specialize services, it can present security issues with different infrastructures, different controls, widely dispersed identities, communication between clouds, and policy enforcement complexity. This conceptual study explores the potential of combining Artificial Intelligence (AI) capabilities with Zero Trust Architecture (ZTA) to enhance adaptive threat detection and prevention in multi-cloud environments. The study is a qualitative exploratory study using documentary and secondary sources, only scholarly literature and cyber security authorities' guidance. Thematic analysis and synthesis of documentary evidence were conducted around five themes: multi-cloud security complexity, Zero Trust and identity-centric security, AI-supported threat detection and adaptive response, AI–Zero Trust integration, governance, explainability, privacy, human oversight and resilience. AI can be used to deliver analytical capabilities for detecting anomalous patterns, interpreting security telemetry, and assisting with contextual risk assessment, while Zero Trust offers principles and mechanisms to ensure continuous verification, least-privilege access, policy enforcement, and resource protection. The resulting framework is a continuous loop that connects multi-cloud telemetry, AI powered analysis, contextual verification, risk-adaptive access, least-privilege enforcement, micro-segmentation, adaptive response and continuous monitoring. The cross cutting safeguards are governance, privacy, explainability, human oversight and resilience. The framework is theoretical and has not been empirically applied or statistically proven, and there is no claim of security improvements being measured. The study offers a conceptual framework to support future experimental and organizational studies in the area of AI-powered Zero Trust security for multi-cloud environments.