Operationalising Responsible AI Governance in Financial Crime Compliance: An Evidence-Based Control Framework for Financial Institutions

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Purpose: This paper analyses how responsible artificial intelligence principles can be translated into auditable controls to support artificial-intelligence-enabled financial crime compliance. Methodology: An integrative literature review and design-science method synthesize the Monetary Authority of Singapore’s Fairness, Ethics, Accountability and Transparency principles, the US National Institute of Standards and Technology Artificial Intelligence Risk Management Framework, Financial Action Task Force guidelines and selected international standards. Findings: The paper designs the Governance to Evidence Responsible AI Framework, which contains six governance control domains: mandate and ownership, data and fairness, model validation, decision orchestration, human accountability and continuous assurance. The transaction monitoring application example shows how every governance requirement can be matched to a right of accountable decision, an operational control and evidence retention. Conclusion: The responsible AI adoption requires controlling the whole decision pathway rather than evaluating model accuracy. The organizations must be able to reconstruct and challenge both decisions to escalate and decisions not to escalate. Practical implications: Financial institutions should deploy use-case classification proportionally to the use-case risk level, independent validation, human oversight that carries meaning, versioned decision logs, closure sampling and suspension triggers. Data Availability Statement: The paper describes a conceptual framework and does not involve any human participants, confidential data about customers and production data.