As AI reshapes financial services at an accelerating pace, are regulatory authorities keeping up — or falling further behind?
The Cambridge Centre for Alternative Finance (CCAF) at the University of Cambridge, in collaboration with the BIS, IMF, World Economic Forum, IDB, CGAP, and the Arab Monetary Fund, has published the 2026 Global AI in Financial Services Report. Drawing on 628 respondents — including 130 financial regulators and supervisors — across 151 jurisdictions, the report offers a comprehensive picture of AI adoption, impact, and risk across the financial ecosystem.
The findings raise urgent questions about the widening supervisory gap. While 40% of financial institutions have reached advanced stages of AI adoption, only 20% of regulators have done so — with 48% still in the Exploring stage or not engaged at all. The gap is even wider in emerging markets, where only 10% of EMDE regulators report advanced adoption versus 33% in advanced economies. Regulators’ top risk concerns include data privacy and protection (80%), model hallucinations (70%), adversarial AI and cyber threats (57%), and critical third-party concentration (43%). A structural challenge looms: 79% of regulators consider explainability critical to their objectives, yet only 50% of supervised institutions have adopted explainable AI methods. The rapid uptake of agentic AI — already at 52% among industry respondents versus 28% for regulators — deepens the supervisory blind spot further. On an optimistic note, 78% of regulators view AI as significant or transformative for their own objectives by 2030, and there is broad convergence across all groups on the urgent need for clearer regulatory guidance.
