Regulators Accelerate AI Adoption: FINMA Chair's Latest Remarks
Marlene Amstad, Chair of the Swiss Financial Market Supervisory Authority (FINMA) and Chair of the International Organization of Securities Commissions (IOSCO), has stated that regulators are accelerating the adoption of artificial intelligence and regulatory technology (RegTech) tools, including those designed specifically for overseeing crypto markets. This statement reflects a broader trend among mainstream global regulators to embrace technology-driven oversight, especially as digital asset markets evolve rapidly and pose new challenges for financial stability and investor protection.
Real-Time Crypto Monitoring Dashboard: Identifying Concentration and Operational Risks
FINMA has already developed and deployed a real-time crypto monitoring dashboard. The dashboard integrates quarterly reports from institutions on their crypto asset holdings with daily market price data. It is designed to identify two critical types of risk: first, excessive concentration of crypto exposure at a single institution; second, operational risk arising from tokens being concentrated on a single blockchain. By aggregating data in real time, supervisors can quickly detect changes in institutional risk exposure amid market volatility and intervene proactively before risks escalate.
Generative AI with Dual-Review Mechanism: Enhancing On-Site Inspection Efficiency
Amstad also revealed that FINMA is developing a generative AI tool. The tool will automatically scan large volumes of regulatory documents before on-site inspections and flag anomalies or potential compliance issues. To ensure accuracy, these flagged items are then independently reviewed by a second AI tool, and the consolidated recommendations are finally submitted to human regulators for judgment. This two-stage mechanism—AI initial screening plus AI cross-check—aims to reduce false positives, improve inspection efficiency, and preserve human oversight for final decision-making. The approach demonstrates how generative AI can be safely integrated into high-stakes regulatory processes with built-in checks.