A Survey of Zero-Trust Architectures for Artificial Intelligence (AI)-Enabled Payroll ERP Security

Abstract

AI-based Enterprise Resource Planning (ERP) payroll solutions boost automation, analytics, and operational efficiency but increase the cyber risk surface many times over by centralizing, integrating, and automating processes with clouds. Traditional security perimeter models currently in use are ineffective at protecting such dynamic environments against insider attacks, fraud, and advanced persistent attacks. Zero-Trust Architecture (ZTA) helps address these issues by providing verification, least-privileged access, micro-segmentation, and policy-based enforcement. The author presents the concepts of zero-trust architectures in this survey, which align with AI-enabled payroll ERP security, including key concepts, identity-conscious controls, behavioral analytics, AI-assisted anomaly detection, and context-dependent access controls. The architectural elements, policy orchestration and edge AI methods of continuous authentication and adaptive authorization. It also contrasts the latest studies and models based on AI and Zero Trust used to detect fraud and ensure safe ERP activities.