A Review of AI-Based Data Governance Frameworks for Enterprise Data Systems

Abstract

Enterprise data governance is changing as Artificial Intelligence (AI) is implemented that enables automated proactive and
scalable management of the data assets throughout their lifecycle. The work showcases how AI can improve modern enterprises’
decision-making, security, compliance, and quality. It offers the postulates of the data governance core, e.g. policy management,
metadata management, and lineage management, as well as privacy, lifecycle control, and explains how AI technologies, e.g.
ML, NLP, anomaly detection, federated learning, differential privacy, and homomorphic encryption are used to improve such
functionality. The study also contrasts the traditional forms of governance systems and those of the AI systems and observes
that they are more precise, scalable, transparent, and real-time responsive. The architecture of AI-based systems of governance is
presented which consists of the ingestion, metadata services, AI analytics, policy enforcement, observability, and access control
layers. The paper also addresses the enterprise frameworks, ethical considerations, and risk management strategies. Finally, it
identifies crucial problems such as the presence of the data silos, the problems of interoperability, and regulatory constraints
and considers the recent developments around the autonomous, privacy-centric, and regulation-based systems of governance.