AI Governance Mechanisms in Modern Financial Risk Management Systems

Abstract

This article investigates how artificial intelligence (AI) is governed within modern financial risk management systems, asking under what conditions AI-based models can enhance, rather than undermine, risk control. The study conducts a systematic literature review of peer-reviewed articles published between 2019 and 2024, focusing on governance mechanisms associated with AI applications in credit, market, liquidity, and operational risk management. The reviewed evidence shows that AI governance is still emerging and uneven, with most institutions extending traditional model risk management frameworks while struggling to address data drift, feedback loops, bias, and systemic effects. The article discusses the literature through a narrative and thematic synthesis that maps governance practices across three main dimensions: AI-specific model risk management, the use of explainable AI as a governance tool, and organizational and ethical mechanisms such as human-in-the-loop oversight and legal accountability. The main findings highlight fragmented implementation, limited empirical evaluation of effectiveness, and the need for more coherent, testable governance architectures.