AI governance and algorithmic auditing in financial institutions: Lessons from Singapore
Abstract
This paper examines the role of algorithmic auditing as a mechanism for responsible AI development and deployment in the financial sector, with a particular focus on Singapore’s regulatory and institutional initiatives. Against the backdrop of fragmented global artificial intelligence (AI) governance frameworks, the study analyses how Singapore has developed operational tools — such as the Veritas Toolkit, AI Verify, Project Moonshot and Project Mindforge — that go beyond abstract ethical principles to provide measurable, use-case-specific standards for auditing AI systems. These initiatives contribute to standardising audit practices, enhancing transparency and bridging trust gaps between financial institutions, regulators and stakeholders. The paper finds that Singapore’s model is notable for its regulator-led, collaborative approach and its focus on sectoral applicability, particularly in high-risk areas such as credit scoring and fraud detection. It also identifies key limitations: the voluntary nature of these frameworks, the challenges of replicability in larger or more fragmented jurisdictions and the lack of universally accepted standards for algorithmic auditing. Moreover, the study highlights the need to broaden auditing efforts to include organisational and human factors, recognising that the use and interpretation of AI outputs are equally critical in managing risk. Ultimately, the paper offers insights into how Singapore’s experience can inform the development of scalable, enforceable and effective algorithmic auditing frameworks in global financial services. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
The full article is available to subscribers to the journal.
Author's Biography
Nydia Remolina is Assistant Professor of Law at the Yong Pung How School of Law, Singapore Management University and Head of Industry Relations at the SMU Centre for Digital Law. She is also a member of the Swiss FinTech Innovation Lab at the University of Zurich. Her main areas of work and academic research include financial regulation, banking law and the intersection of law, technology and finance. Nydia has taught or delivered lectures at various institutions and programmes across the US, Asia, Europe and Latin America, including the Global Certificate Program jointly organised by Harvard Law School and the International Organization of Securities Commissions (IOSCO), and the course on open finance offered to financial regulators and policy makers by the Centre for Alternative Finance at the University of Cambridge. Prior to joining academia, Nydia practised law, serving as the Manager of Policy Affairs at one of the largest financial conglomerates in Latin America, as a Senior Adviser to the Organization for Economic Cooperation and Development (OECD) and working at Sullivan & Cromwell’s New York Office.
Citation
Remolina, Nydia (2025, June 1). AI governance and algorithmic auditing in financial institutions: Lessons from Singapore. In the Journal of Risk Management in Financial Institutions, Volume 18, Issue 3. https://doi.org/10.69554/RVPW6910.
Publications LLP