Governance Frameworks for Enterprise AI Systems operating in Regulated Environments

Enterprise AI systems are being deployed at unprecedented speed across highly regulated sectors, yet governance frameworks have not evolved fast enough to prevent systemic risk, compliance failures, and opaque decision-making.As organizations increasingly rely on complex architectures, including generative AI, agentic systems, and distributed multicloud pipelines, traditional governance models built for deterministic IT systems are no longer fit for purpose.This study addresses this critical gap by conducting a systematic literature review of emerging governance studies published between 2024 and 2025, a period defined by the rollout of the EU AI Act and the global rise of enterprise-scale AI adoption.Drawing on evidence from contemporary scholarship, the study proposes a five-layer Enterprise AI Governance Framework that integrates strategic governance, lifecycle and operational oversight, autonomous system control, explainability and human oversight, and data and infrastructure governance.The synthesis reveals that while data governance and cybersecurity practices are relatively mature, significant weaknesses persist in strategic alignment, continuous lifecycle governance, and the oversight of autonomous and agentic AI systems.Explainability remains inconsistently implemented despite regulatory mandates, and organizations struggle to operationalize human-in-the-loop mechanisms at scale.The study contributes a novel, integrated governance architecture grounded in empirical literature, as well as an extended governance matrix and operationalized constructs that translate abstract principles into actionable controls.The findings highlight the urgent need for coordinated, multi-layer governance capable of addressing crossorganizational, cross-regulatory, and cross-lifecycle risks.This research provides a timely foundation for strengthening accountability, transparency, and compliance in enterprise AI systems operating in rapidly evolving regulatory environments.