Observation

How to keep SupTech future-proof

Denis Beau Intervention

Denis Beau, First Deputy Governor of the Banque de France

Published on 20 September 2023

Denis Beau Intervention

Supervision Innovators Conference, 20 September 2023: “How to keep suptech future-proof”

Prepared initial remarks by Denis Beau, First deputy-governor of Banque de France.

The digitalization of finance, the emergence of new players and technologies as well as associated risks, including cyber risks: all these trends make it urgent to develop an enhanced approach to financial supervision.

How do we meet this challenge, at the ACPR and Banque de France?

I. First, by adopting a proactive and demand-driven approach to innovationwhich has resulted in the launch of a "suptech" programme in 2019This programme aims to strengthen the ACPR's supervisory capabilities using new data processing and artificial intelligence technologies.

A/ All the projects identified, developed and deployed since 2019 are the result of a structured process based on a number of incentives and filtersI would like to mention two in particular, which play a very important role.

One very important incentive is provided by the strong impulse and involvement in the selection and monitoring of the investigation, implementation and deployment phases of the projects from the highest levels of the management chainwho set “the tone from the top”. This strong impulse and involvement helps to overcome the inevitable - and understandable - resistance generated by the transformation of established business processes. To ensure that this momentum is maintained overtime, the suptech programme has been included in the Banque de France's 2019-2024 strategic plan, and features prominently in the list of priority actions monitored by the Executive Committee.

Among the filters used, one of the most important is that the projects contribute to solving the day-to-day operational problems faced by ACPR staff in the course of their duties. In other words, our approach is demand-driven, rather than supply-drivento avoid the classic situation of solutions in search of a problemwhich end up sitting on the shelf. To do so, as we know that it is hard to imagine an innovative solution without knowing what technology can and cannot do, we now have supervisory experts and IT staff working together in project-oriented teams combining creativity and know-how.

In addition to these incentives and filters, the top management has a role to play to promote more efficient and less costly approacheswith attention paid to the results of each project. Being more efficient means giving priority to agile project management methodswhich allow issues to be corrected as they arise, thereby increasing productivity. It also means monitoring the project from end to endright through to the industrialisation phases. Finally, it means encouraging feedback on the difficulties encountered during a project, so that we can make collective progress on subsequent projects.

BThis approach feeds into and forms part of an overall reflection on the evolution of IT systemsfocused on the question of data and its use.

Whether structured or unstructured, data is the key element in the ACPR’s analyses and controls, and it is growing exponentially. The suptech projects all aim to facilitate the management, access and visualisation of this data, to enable better targeted analyses and controls to support our risk-based supervision and to improve its effectiveness and efficiency.

Moreover, the suptech approach cannot afford to overlook the technical problems linked to the management, evolution or decommissioning of existing systemsThe insertion of new tools into the existing environment cannot be improvised, but must instead be designed in a way that is consistent with developments in the information system as a whole. In addition, decommissioning old applications - which may be one of the objectives of deploying new tools - should also be integrated into the data enhancement strategy: in many cases, legacy applications contain a wealth of data, which should not be overlooked.

II. To be effective, this approach must also be carried out in a spirit of anticipation and openness.

A/ We must first anticipate the profound changes in our practices and organizations that some new technologies are likely to bring about.

For example, generative AI offers exciting opportunities, particularly in terms of synthesising, extracting, and classifying information, or generating text or computer code. At the same time, it raises a large number of concerns, in terms of reliability, data confidentiality or intellectual property.

To anticipate the effects of these new technologies, we need to go through the learning curve as quickly as possible, to get as complete a picture as possible of their potential benefits, but also of their risks. In particular, the experimentation approach can facilitate accelerated learning.

For example, in May 2023, the Banque de France, through the innovation platform operated by its Lab, launched a call for contributions on the uses and impacts of generative AI on its activities. Addressed to the innovative ecosystem, it aims, on the one hand, to refine our understanding of this technology and, on the other hand, to accelerate its potential use, through the proposal of concrete use cases or even operational solutions.

Anticipation must also concern the human resources needed to adapt to new modes of supervision. This includes diversifying the experience and profiles of supervisors, by fostering the recruitment and development of specialized skills, notably in data science and artificial intelligence.

B/ We must also be open-minded because, in an area where solutions have yet to be invented, cooperation and co-construction can bring many shared benefits.

As we have seen, cooperation with the innovative ecosystem can be very fruitful; as well, we have a strong interest in maintaining links with the academic community. Here, however, I would like to put particular emphasis on the value of establishing enhanced cooperation with other authorities (MoUs, multi-authority supervisory committees, etc.), in order to develop shared analyses of topics of common interest (e.g. the benefits and risks of a new technology) or to jointly develop tools useful for the supervision of the financial sector.

The Single Supervisory Mechanism (SSM)the forum par excellence for cooperation between national and European supervisory authorities, is the ideal setting for organizing such cooperative mechanisms. In this regard, the ACPR and the Banque de France are fully committed to providing all European partners with a “suptech centre” that will offer, for example, an advanced analysis interface for supervisors, capable of integrating analysis modules developed by other authorities or SSM participants.

Updated on 21 November 2023