As an extension of its work on artificial intelligence (AI), and the publication of two discussion papers, the ACPR wants to organize an experiment to evaluate the benefits of combining data on model performance of detecting suspicious transactions in the area of LCB. -FT in the banking sector.
According to the Financial Action Task Force (FATF), which published a report on this topic in 2021, with data pooling and collaborative analysis, financial institutions can better understand, assess and reduce the risks of money laundering, capital and financing of terrorism and so more easily. identify illegal activities.
Due to the innovative nature of its format and the technologies used, this experiment will take place in several stages:
- A presentation meeting on March 30, 2022 for players interested in participating in this experimental project: banks and technical service providers.
- A “Tech Sprint”, enabling technical service providers to provide solutions for collaborative computing and data pooling
- to present their expertise to willing banks.
- In parallel, reflection and modeling workshops with the banks participating in the experiment, in particular to establish the most interesting use cases to test.
The implementation of the selected methods and the evaluation of the results by each of the teams made up of banks and service providers.
solutions based in particular on “PET” (Privacy-enhancing technologies or methods for the preservation of confidentiality) With this experiment, the objective of the ACPR is to stimulate the market’s reflection on the applications of the AI in the field of LCB-FT. More specifically, this experiment will be an opportunity, testing solutions for pooling or collaborative data analysis, to explore ways to improve the detection algorithms, likely to reduce the number of false alerts and thus increase the effectiveness of the fight against money laundering.