Thematic review on the use of Artificial Intelligence in the Luxembourg Financial sector

The Thematic Review on the Use of Artificial Intelligence in the Luxembourg Financial Sector report summarizes the findings of a joint survey conducted by the BCL and the CSSF between October 2021 and January 2022.
The survey aimed to understand the level of adoption of innovative technologies, particularly AI and ML in the Luxembourg financial sector. The survey was sent to all credit institutions, payment institutions, and e-money institutions supervised by the CSSF, and 138 out of 148 institutions participated in the survey.
The report is divided into five parts. The first part provides information about the survey demographics, including the type of entities and the profile of survey contact persons. The second part focuses on the digital strategy of the institutions, including their investments in innovative technologies such as AI, APIs, digital onboarding techniques, and DLT. The third part covers the adoption of AI and ML technologies, including benefits and challenges, data science team organization, data governance, security and robustness, machine learning development lifecycle, and technical infrastructure. The fourth part focuses on the specific use cases of AI technology, including use case categories, general aspects, AI technologies, type of ML problems, type of learning, and ML algorithms. Finally, the fifth part discusses AI trustworthiness, including human in the loop, bias, auditability, explainability, and security testing.
The report found that the level of adoption of AI and other innovative technologies was fairly limited and still at an early stage. The type of innovative technology that received the most adoption was APIs, followed by digital onboarding and then AI, while only a small percentage of respondents invested in DLT. However, the responses indicate a general increase of investments across all categories of innovative technologies compared to the 2021 budget, with the highest increase for ML technology. The report found that there is a low level of maturity regarding the implementation of ethical control measures such as those related to fairness and bias.