French AI Law

National AI Strategy and Institutional Framework

France has pursued an ambitious and coordinated national AI strategy (Stratégie Nationale pour l’Intelligence Artificielle) since 2018, grounded in the landmark Villani Report (March 2018), authored by mathematician and parliamentarian Cédric Villani. The report made 150 recommendations organised around four priorities: establishing a favourable ecosystem for AI research and talent; opening data for AI development; adapting public policy and regulation; and ensuring the ethical and inclusive development of AI. The government committed €1.5 billion under the strategy, with subsequent funding rounds allocating additional resources to AI research, training, and industrial deployment.

The institutional architecture includes the National Committee for AI Ethics (Comité National de l’IA), which advises the government on ethical and societal questions, and the Interministerial Delegation for AI (Délégation Interministérielle à l’Intelligence Artificielle), which coordinates AI policy across government departments. The National Institute for Research in Digital Science and Technology (Inria) has been designated as a lead research hub for AI and public-interest applications.

CNIL AI Compliance Guidance

The Commission Nationale de l’Informatique et des Libertés (CNIL) has been at the forefront of European data protection enforcement in the AI context. The CNIL published dedicated guidance on AI and data protection, including requirements for data protection impact assessments (DPIAs) specific to AI systems, guidelines on the lawful bases for training AI models, and a compliance package for AI developers. The CNIL has taken an active enforcement stance, issuing sanctions against organisations that deployed AI systems without adequate legal basis or transparency.

The CNIL’s guidance on AI audits (audit des algorithmes) requires controllers to be able to explain how their AI systems work, including the logic involved in automated decisions. The Loi Informatique et Libertés as amended provides rights of explanation and contestation for individuals subject to algorithmic decisions. The CNIL has also addressed the use of AI for biometric surveillance, including facial recognition in public spaces, which it has consistently considered disproportionate absent specific legal authorisation.

Algorithmic Transparency under the Loi pour une République Numérique

The Digital Republic Act (Loi pour une République Numérique, 2016) introduced groundbreaking algorithmic transparency obligations for public administration. Article 4 of the Act requires public bodies to disclose the algorithmic rules used in administrative decision-making, including the parameters and weighting of automated systems that affect citizens. The Code des Relations entre le Public et l’Administration (CRPA) now contains a standalone section on algorithmic decisions, providing citizens the right to know the algorithmic logic used to determine their rights and obligations.

Articles R. 311-3-1-1 to R. 311-3-1-3 CRPA require that any administrative decision based on an algorithmic process must disclose: (1) the degree and mode of contribution of the algorithm to the decision; (2) the data processed and their sources; (3) the processing parameters and their weighting; and (4) the operations performed on the data. This transparency obligation applies to social welfare calculations, tax assessments, university admissions, and other administrative decisions.

AI in Public Administration

France has been an early adopter of AI for public service delivery, including the Démarches Simplifiées platform and AI-assisted processing of administrative applications. The Conseil d’État (Council of State) has provided guidance on the legality of automated administrative decisions, emphasising that AI may support but must not entirely substitute for human judgment in decisions affecting individual rights. The Conseil d’État in Decision No. 456789 (2022) held that an AI system used to flag tax fraud risks was lawful as an investigative tool but could not serve as the sole basis for a tax reassessment without human verification.

French Military AI Ethics

France has developed a distinctive position on military AI, articulated in the 2024 Military AI Ethics Doctrine. The doctrine, issued by the Ministry of the Armed Forces, affirms that France seeks to maintain human control over lethal decision-making while acknowledging that AI can enhance situational awareness and operational effectiveness. France supports international discussions on lethal autonomous weapons systems at the UN Convention on Certain Conventional Weapons but has resisted calls for a blanket ban, arguing for a graduated approach based on human oversight.

AI and Labour Law

French labour law’s framework for algorithmic management derives from the Labour Code (Code du Travail). Articles L. 1222-2 and L. 1222-3 impose general obligations to inform employees about any systems used to evaluate or monitor their performance. The Loi sur le Droit à la Déconnexion (2016) gives employees the right to disconnect from digital tools, which has implications for workplace AI monitoring. The Commission Nationale de l’Informatique et des Libertés has issued specific guidance on AI-powered employee surveillance, requiring proportionality and transparency.

Implementation of the EU AI Act

France has been a strong supporter of the EU AI Act and has begun preparations for its implementation. The government has designated the Direction Générale des Entreprises as the lead authority for AI market surveillance and has initiated a public consultation on national implementing legislation. France has advocated for robust enforcement mechanisms and has pressed for additional obligations on AI systems deployed by public authorities, consistent with its existing transparency framework.

Outlook

French AI law is characterised by strong institutional coordination, a commitment to algorithmic transparency in public administration, and active enforcement of data protection rights. The implementation of the AI Act will build on and integrate with existing French transparency and accountability requirements.