Regulatory framework for artificial intelligence (AI) in the financial sector: lessons from the European Union AI act

Purpose The paper aims to develop a comprehensive regulatory framework for the use of Artificial Intelligence (AI) in the financial sector. Drawing on the European Union’s Artificial Intelligence Act (EU AIA), the study develops a risk-based proportional approach to AI regulations applicable across financial markets and institutions. Design/methodology/approach The paper employs a qualitative research design using a two-phase analytical approach. Firstly, it identifies specific AI risks and assesses their impact on financial regulatory objectives of financial stability, consumer protection and financial integrity. Secondly, it applies a risk-based proportional regulatory approach, drawing on the EU AIA to outline specific mechanisms for risk mitigation, governance and oversight to regulate AI risks. Findings The study finds that AI introduces new layers of risk and transmission channels that can affect financial markets and institutions. The paper applies the framework of EU AIA for classifying AI applications by risk levels (unacceptable, high, limited, minimal) and identifies corresponding tailored regulatory measures. Effective implementation of AI regulations depends on integrating these measures into strong internal governance and risk management frameworks within financial institutions. Practical implications The paper provides a risk-based regulatory design for AI governance for financial institutions by systematically applying the approach of the EU AIA and offers guidance for policymakers developing sector-specific AI oversight frameworks. Originality/value To the best of the author’s knowledge, this paper provides a novel contribution by offering one of the first structured frameworks for regulating AI in the financial sector, an area with scant literature.