Abstract

The rapid integration of artificial intelligence (AI) into financial systems has transformed core financial functions, including credit allocation, fraud detection, algorithmic trading, and regulatory compliance. While AI-driven financial technologies promise enhanced efficiency, predictive accuracy, and financial inclusion, they also introduce significant ethical, legal, and systemic risks that challenge existing governance structures. These risks ranging from algorithmic discrimination and opacity to accountability gaps, privacy violations, and threats to financial stability are amplified by the scale, interconnectedness, and high-stakes nature of financial decision-making. Current ethical AI frameworks and regulatory responses, although valuable, often rely on principle-based or uniform governance approaches that fail to account for the heterogeneous risk profiles of financial AI applications. Moreover, compliance-oriented regulatory regimes typically establish minimum standards and may lag behind technological developments, limiting their effectiveness in managing emerging ethical risks in real time. This paper advances a finance-specific, risk-based framework for ethical AI governance that aligns the intensity of oversight with the likelihood and severity of potential harm. By embedding ethical considerations within established financial risk management and operational resilience practices, the proposed framework provides a structured and scalable approach to identifying, assessing, and mitigating ethical risks across the AI lifecycle. The framework emphasizes proportionality, accountability, transparency, and continuous monitoring, addressing both individual-level harms and system-wide stability concerns. Ultimately, the study argues that ethical AI governance in finance must move beyond compliance toward responsible innovation that sustains trust, resilience, and legitimacy in increasingly AI-driven financial systems.