FRANCE Trends and Developments Contributed by: Eric Weil, Victor Gozzerino and Giada Cancemi, Weil & Associés
French courts may thus scrutinise more closely the behaviour of victims of AI-facilitated fraud, and in par - ticular the measures deployed to prevent and detect this risk, measures which may themselves increas - ingly be developed with the assistance of AI. AI as a tool for fraud detection and prevention AI is not solely an accelerator of fraud. It is also becoming a tool for detection and prevention. In France, this evolution is particularly evident in the field of anti-money laundering and counter-terrorist financing (AML/CFT). Article L. 561-2 of the Monetary and Financial Code imposes obligations on a broad range of entities, including financial and insurance sector institutions, as well as certain regulated pro - fessions. Financial sector actors are required to imple - ment customer due diligence measures, compelling them to gather information on their clients in order to understand the risks each client presents and the transactions they are likely to undertake, to build a client profile and thereafter to monitor, on an ongoing basis, that transactions correspond to that profile. In this context, the joint guidelines issued by the ACPR ( Autorité de contrôle prudentiel et de résolution ) and Tracfin ( Traitement du renseignement et action contre les circuits financiers clandestins ), published on 23 April 2025 and replacing the guidance in force since 2018, recommend the use of AI to refine client segmentation, improve alert mechanisms, and detect money-laundering networks or circuits that are not identifiable through traditional approaches. The primary contribution of AI lies in the calibration of alerts. Where a conventional system triggers an alert upon the crossing of a fixed threshold, a more advanced system can assess a transaction by ref - erence to the client’s actual profile, sector, turnover, payment habits or a comparable peer group. Machine learning and AI can detect atypical behaviour, reveal new risk typologies and adapt alert-triggering criteria to evolving client behaviour. Human oversight nevertheless remains central. AI can - not, on its own, form the basis for a decision to submit a suspicious activity report to Tracfin. Under French law, such a report requires an individualised assess -
ment of the facts, characterisation of the suspicion, and adequate documentation, notably in accordance with Articles L. 561-15 and R. 561-31 of the Monetary and Financial Code. European supervision and compliance evidence These developments are taking place against a more significant European backdrop. With effect from 10 July 2027, the European Anti-Money Laundering Reg - ulation (AMLR) will be directly applicable throughout the European Union, establishing a more uniform body of rules for obliged entities. For financial groups operating in multiple member states, the change is material: customer due dili - gence, know-your-customer, risk classification and monitoring obligations will no longer be assessed solely through the lens of potentially divergent national practices, but within a more harmonised European framework. It is here that AI can prove genuinely useful, not because the new European framework mandates it, but because a more harmonised AML/CFT regime requires the processing of significant volumes of data, the identification of inconsistencies, the prioritisation of alerts and the documentation of decisions made by compliance teams. Commercial communications and regulated professions A further area of concern involves commercial com - munications. A company that uses AI to generate advertising, investment promises, commercial docu - ments or presentations of financial services cannot shelter behind automation where the message mis - leads the public. The position becomes yet more sensitive with respect to generative tools presenting themselves as lawyers. Under French law, Article 433-17 of the Criminal Code criminalises the unlawful use of a title associated with a regulated profession. Articles 54 and 66 of the Act of 31 December 1971 further regulate the provision of legal advice and the drafting of legal documents on behalf of third parties. An automated service providing individualised legal
66 CHAMBERS.COM
Powered by FlippingBook