Using AI for financial supervision

Financial intermediaries use Artificial Intelligence (AI) applications for purposes ranging from the detection of mistrades and fraud to the optimization of market-making and hedging strategies. In addition, AI is increasingly relied upon for financial statement analysis, the detection of accounting misstatements, and securities litigation.

Prof. Gerard Hertig, from the Future Resilient System (FRS) programme, identifies the quantitative and qualitative importance of AI for the provision and supervision of financial services. His analysis is based on a review of current academic contributions and supervisory practices.

As for many innovations, the rise of AI occurred in a regulatory vacuum. Consequently, there is an ongoing race to design regulatory frameworks that are both AI-friendly and safe. Simple rules are likely to remain the fundamental guideline for the design of legal institutions; however, AI developments are expected to significantly reduce lawmaking and enforcement costs.

For their part, supervisory authorities are increasingly relying on AI for internal purposes. On the other hand, their use of AI for regulatory purposes remains mostly confined to detecting illegal market practices with the exception of the US, where States are the main players and the regulatory scope is broader than in the rest of the world.

For full details, please download the working paper Download “Using artificial intelligence for financial services and supervision” (PDF, 469 KB).