Artificial Intelligence law at Réunion (France)
1. Applicability of AI Law to Réunion
Réunion is an overseas department and region (DROM) of France.
Under the French Constitution (Article 73), national laws apply automatically in Réunion, unless a specific law provides a local adaptation.
Therefore, all AI-related laws, regulations, and judicial decisions that apply in mainland France also apply in Réunion.
Implication: There is no separate AI law for Réunion; it is governed by French and EU law.
2. Legal Framework Governing AI
AI in France (and Réunion) is not regulated by a single law but through a combination of existing legal regimes:
| Area | Key Rules for AI |
|---|---|
| Data Protection / Privacy | AI systems processing personal data must comply with data protection principles: lawfulness, transparency, purpose limitation, and protection of sensitive data. |
| Criminal Law | Misuse of AI, such as non-consensual deepfake distribution, can result in criminal liability. Penalties include fines and imprisonment. |
| Labour / Employment Law | Introducing AI in the workplace requires consulting employee representatives or works councils. Failure to do so can result in injunctions or penalties. |
| Intellectual Property / Copyright | Training AI on copyrighted works may require authorization from rights holders. Unauthorized use can trigger infringement claims. |
| Sector-Specific Rules | Certain sectors, like public surveillance or medical AI, have additional legal requirements, including authorization, transparency, and safety obligations. |
Note: EU law, including the proposed AI Act, heavily influences French regulations and will apply to Réunion as part of the EU legal framework.
3. Key Case Law
A. AI in the Workplace
Case: Nanterre Court of Justice, 2025
Facts: A company deployed AI software in the workplace during a “pilot phase” without consulting the works council.
Ruling: The court ordered suspension of AI deployment. It emphasized that even experimental deployment triggers labour-law obligations, such as consultation with employee representatives.
Implication: Employers in Réunion must comply with workplace consultation requirements before deploying AI systems.
B. AI-Generated Content and Copyright
Case: SNE / SGDL / SNAC v. Meta, 2025 (pending)
Facts: Publishers sued Meta, alleging unauthorized use of copyrighted books to train AI models.
Legal Issue: Whether training AI on copyrighted material without consent constitutes infringement.
Implication: Could set a precedent for AI training datasets in Réunion, requiring authorization for copyrighted works.
C. Criminal Liability for Non-Consensual AI Use
Legal Principle: Non-consensual creation or distribution of AI-generated images or videos (deepfakes) of identifiable individuals is a criminal offence in France.
Penalties: Fines, imprisonment, and enhanced penalties for sexualized or harmful deepfakes.
Implication: Individuals and companies in Réunion face criminal liability for misuse of AI-generated content.
D. AI and Public Surveillance
Legal Principle: Deployment of AI in public surveillance requires authorization, transparency, proportionality, and compliance with data protection.
Implication: Public authorities in Réunion using AI-based cameras or monitoring tools must comply with national laws to avoid legal challenges.
4. Key Legal Principles
Procedural Compliance is Mandatory: Whether for employment consultation, copyright licensing, or AI deployment in public spaces, legal procedures must be followed.
AI Misuse Can Trigger Criminal and Civil Liability: Unauthorized or harmful uses of AI can lead to prosecution or civil claims.
IP and Data Protection Are Central: AI training, operation, and deployment must respect copyright, licensing, and privacy rules.
French Law Applies Directly to Réunion: All case law and regulations from mainland France are binding, except where local adaptations are explicitly provided.
5. Practical Implications for Réunion
Employers must consult works councils before implementing AI systems.
AI developers must ensure data protection compliance and secure rights for any copyrighted material used.
Individuals and organizations must avoid non-consensual AI-generated content, particularly images or videos.
Public authorities must justify and regulate AI surveillance to comply with transparency and proportionality requirements.
Ongoing litigation in France may further clarify liability and obligations, and these developments will apply to Réunion as well.
Summary Table
| Area | Rules / Principles | Relevant Case / Example |
|---|---|---|
| Workplace AI | Consultation required | Nanterre Court, 2025 |
| AI-generated content | Copyright compliance | SNE v. Meta, 2025 |
| Criminal use | Deepfake restrictions | French Penal Code (amended for AI) |
| Public surveillance | Authorization, transparency | Sector-specific regulations |
Conclusion:
Réunion is fully subject to French and EU AI laws. The main focus areas include workplace deployment, copyright compliance, criminal liability, and public surveillance. Courts have already applied these principles in France, creating precedents that automatically affect Réunion. Companies and authorities must carefully follow these rules to avoid legal risks.
RELATED Blog
comments