Next content

Read more


The pandemic, Government-to-person (G2P) programs, and digital wallets

The SARS-CoV-2 pandemic, an exogenous, sudden, and symmetric shock, has provoked unprecedented challenges in most of the world by halting the real economy. The combination of global chain disruptions affecting the supply, and containment...

Artificial Intelligence (AI) and Machine Learning (ML) have become an increasingly integral part of central banks’ and credit institutions’ daily work and of financial market operations. Banks, insurance companies, financial regulators and supervisors constantly deal with and have access to an incredibly large amount of data. These technologies not only contribute to process and analyse big data, but algorithms are also capable to infer and predict tendencies and detect potential risks to financial stability – among others.  While the most common and known uses of Big Data in the sector, AI and ML, are connected to financial services and banking products, there are also important applications of these technologies in many areas relating to financial regulation and supervision. There is a large list of applications of AI and ML in finance including risk assessments, macroeconomic analysis, algorithmic trading, forecasting and monitoring of financial market indicators.

Since the aftermath of the global financial crisis, regulation and reporting requirements contributed to significantly increase the amount data reported to central banks and supervisory authorities. AI and ML tools are also being more frequently used for supervision and regulatory reporting (i.e.. Suptech and Regtech), making it easier for supervisors and regulators to collect, store, analyse and assess large amounts of data more efficiently. However, supervisors and regulators also experience challenges due to the fast-changing digital technological environment. Likewise, as AI and ML are transforming the financial sector, they are also bringing some legal uncertainty and data privacy challenges for central banks and supervisors, which are mainly associated to data quality, consumer protection, data ownership and privacy problems, transparency issues and ethical considerations, among others.

AI and ML will continue to be a major trend in the future as digital innovation evolves. Therefore, financial institutions, regulators and supervisors require to constantly adapt to these digital improvements, specially by adopting and apprehending new models that collect and analyse huge amounts of data and detect anomalies in the financial markets. Against this background, there is an increasing need for supervisors to enhance their knowledge and expertise in the area of digital finance in order to be able to effectively perform their supervisory tasks. The latter also requires the creation of a “digital culture” in the supervisory and regulatory spheres. This is not an easy task, but it can be done by diminishing sectoral barriers, especially between academia, fintech companies, policy makers and financial regulators and supervisors, to share knowledge and practices on the use, benefits and risks of these technologies, while at the same time tackling the biased conception of the “incomprehensibility of machine learning tools”.

The Florence School of Banking and Finance (Robert Schuman Centre, European University Institute) organised a workshop with the aim of promoting cross-sectoral dialogue and discussions on the applications and tools of AI and ML in finance, particularly in the context of financial supervision. It also focused on the identification of current trends, risks and challenges relating to the use of these technologies, as well as on the opportunities for regulators and supervisors. The event gathered speakers and discussants from different sectors and fields, including the European Supervisory Authorities (EBA, EIOPA and ESMA), academics, central banks’ representatives, consultants, European Commission’s officers, members of international organisations and national regulators, among others- to exchange views and experiences relating to these timely relevant topics. The programme and list of participants is available here.

In line with the EU 2020 Digital Finance Strategy, and together with the European Commission (DG REFORM) and the three ESAs, the FBF has embarked on the task of contributing to the work of enhancing the supervision of digital finance in the EU and fostering supervisory convergence, through the launch of the EU Supervisory Digital Finance Academy (EU-SDFA). As part of the EC’s Technical Support instrument and in response to the need to enhance the technical knowledge and skills of supervisors and to develop capacity to exploit the potential and impact of technology for supervision, the EU-SDFA will train more than 200 participants from 26 National Supervisory Authorities established in 20 EU member states, covering the entire supervisory landscape of the European financial sector. The EU-SDFA will contribute not only to the development of supervisory tools and practices, but also to the creation of a community of experts at EU level, to the opening of spaces for cross-sectoral dialogue and knowledge exchange, thus reconciling policy, technical and academic perspectives and practices on digital finance – including the use of AI and ML tools. Above all, the EU-SDFA will also aim to establish and foster peer-to-peer synergies and inform policy making at national and EU level.

Back to top