The French AI regulatory landscape
France stands out in Europe for its central position in AI governance: it is preparing a multi-regulator framework expected to be coordinated by the DGCCRF (the formal designation remained unresolved into 2026), participates actively in the work of the European Artificial Intelligence Board, and has developed an ambitious national doctrine with France 2030. That position creates both responsibilities and opportunities for French companies.
French companies face several layers of obligations: the European framework (the AI Act, the GDPR, the NIS2 Directive), the recommendations and decisions of the French authorities (CNIL, DGCCRF, AMF, ACPR, ANSSI), and the case law of the Conseil d'État on algorithms in public services.
The CNIL and AI: the state of the doctrine in 2026
The CNIL has developed a substantial body of doctrine on AI, organised around several strands. On recommender systems, the CNIL requires enhanced transparency and an effective right to object. On biometrics, strict rules apply to facial recognition and behavioural analysis systems. On automated decisions, Article 22 of the GDPR creates rights to explanation and objection that companies must implement in concrete terms.
The CNIL has also published a practical guide on AI and data protection that serves as the operational reference for French DPOs and legal teams. This guide can be relied on against an organisation in the course of a CNIL inspection.
The financial sector: AMF and ACPR
The Autorité des Marchés Financiers (AMF) has published guidance on the use of AI in asset management, financial analysis and algorithmic trading. That guidance creates concrete expectations on model governance, bias testing and the documentation of the AI systems used in investment decisions.
The Autorité de Contrôle Prudentiel et de Résolution (ACPR) supervises AI in the banking and insurance sector. The ACPR has conducted several thematic reviews on AI and has published expectations on the risk management of ML/AI models, notably for credit, fraud detection and insurance pricing.
