In September 2020, the European Commission adopted the digital finance strategy to foster a more competitive and innovative European financial sector. The digital finance platform initiative is part of this effort to support innovation in finance and build a true single market for digital financial services. This platform was created to develop closer relationships between financial firms and supervisory authorities. The platform is a collaborative space that offers practical tools designed to facilitate the scaling up of innovative firms across Member States. It features the new data hub, cross-border services, a fintech mapping, an overview of the latest policy news, calls to action and digital finance events.
The data hub
As you may already know, the European Commission recently launched the data hub – a key novelty of the digital finance platform. The objective is to facilitate the exchange of data between supervisory authorities and financial companies. Commissioner McGuinness officially announced the launch of the data hub at an online event on 21 March 2024.
The data hub provides synthetic supervisory data to financial companies for the purpose of testing new solutions and training artificial intelligence and machine learning (AI/ML) models. The data hub will make it easier to develop products that depend on data-intensive AI systems. On one side, supervisory authorities like national central banks will provide specific datasets that will be synthetized and uploaded on the data hub. On the other side, financial companies will be able to request access to these datasets. Financial companies will therefore have access to large amounts of data that they would not usually have access to, and supervisory authorities will be able to closely follow the development of financial innovations on the EU market. The data hub will therefore complement national sandboxes and innovation hubs that typically focus on facilitating the dialogue between regulators and innovators.
The latest update is that the first datasets have been synthetized and published on the data hub. These datasets were provided by the Central Bank of Lithuania and are related to the central credit register, which records the information lenders submit about loans over €500. At this stage, we consider that this data could be used to train artificial intelligence and machine learning models evaluating the maturities, pricing and amounts of loans being issued to companies. The datasets could also be used to test and validate existing AI models and make sure they perform consistently across diverse scenarios. Another potential use case is the creation of a benchmarking tool that compares data from a credit bureau (a data collection agency) for individual companies with the overall system data. Lastly, the datasets could be used to study historical data patterns and correlations between macro data and possible evolutions of the market.
Synthetic data
To ensure compliance with confidentiality requirements, we have decided to build the data hub using synthetic data. Synthetic data offers a way for supervisors to participate in the project without having to make the real data they hold accessible to any third party. In short, synthetic data ensures full anonymization while preserving the characteristics of the original data that make it relevant for testing purposes. At no point will any real data leave the premises of the respective supervisors, nor will any external user gain access to it.
The synthesis methodology was tested by the joint research centre (JRC), the research and innovation directorate general of the European Commission. The purpose of this testing was to ensure that the new datasets maintain the properties of the original datasets while also protecting privacy and confidentiality.
Call for proposals
To ensure that the data hub offers a clear added value, it is essential to define as precisely as possible which use cases and datasets are most useful for market participants. The European Commission therefore launched a call for proposals in March 2023 to consult a wide range of innovative firms. The objective was to collect information on what type of data financial firms would need to test new applications and train AI models. Most proposals received were related to creditworthiness and fraud detection use cases. In March 2024, we published another call for proposals to continue collecting information on what the market needs in terms of data. The call will be open until 26 April 2024.
Who is involved?
This initiative is steered by DG FISMA’s digital finance unit, in collaboration with the European supervisory authorities (ESAs). DG FISMA is also working with an informal group of supervisory authorities that will provide the synthetic data. Many innovative firms have committed to participating in the development of the data hub by providing feedback on their specific needs in terms of data. DG FISMA is also collaborating with JRC on the synthesis methodology.
If you are interested in learning more, feel free to visit the data hub website and to reach out to eu-digital-finance-platform
ec [dot] europa [dot] eu (eu-digital-finance-platform[at]ec[dot]europa[dot]eu).
Related links
Details
- Publication date
- 10 April 2024
- Author
- Directorate-General for Financial Stability, Financial Services and Capital Markets Union