FAQ – Artificial Intelligence

Artificial intelligence case law

Case law relating to artificial intelligence is currently under construction. In France, there is not yet a body of decisions comparable to that existing in the fields of medical liability, consumer law, or construction law. The initial decisions primarily concern five themes.

1. Use of AI in legal proceedings

Since late 2025, several French courts have noted that legal briefs or submissions contained non-existent legal references (AI «hallucinations»). The courts have refused to take these erroneous references into account and have reiterated that lawyers and parties remain solely responsible for the content of their pleadings.  

2. User Responsibility

The jurisprudential trend is clear: when an AI is used to draft a document, the responsibility remains with its user.

In other words:

  • AI is not a legal person; ;
  • she cannot be convicted; ;
  • liability lies with the professional or individual who uses it.

This principle applies in particular to lawyers, experts, notaries, doctors, and businesses.  

3. AI and Judicial Expertise

A German decision from 2025, widely commented upon in France, held that a court-appointed expert who relied excessively on AI to carry out their mission had failed in their personal obligations as an expert. This case fuels reflection on the duties of experts listed on the court of appeal rosters.  

4. Copyright

The initial litigation primarily concerns:

  • training AI models on copyrighted works; ;
  • the protection of AI-generated content ;
  • the existence (or not) of sufficient human creation.

For example, a German court refused to grant copyright protection to AI-generated logos in the absence of a predominant human creative contribution.  

5. AI in Business

French rulings have considered that the introduction of AI tools could modify employees' working conditions. They have required certain companies to suspend the deployment of these tools until employee representatives had been consulted in accordance with labor law.  

The main principles that emerge

At this stage, case law tends to hold that:

  • AI is a tool, not a legal subject; ;
  • the user remains responsible for the results produced; ;
  • human decision-making remains essential, particularly in the legal, medical, and judicial fields; ;
  • Errors generated by an AI do not constitute grounds for exemption from liability.

Expected trends

In the coming years, litigation is expected to focus primarily on:

  • civil liability in the event of an AI error; ;
  • the responsibility of AI software publishers; ;
  • the protection of personal data; ;
  • copyright ;
  • algorithmic discrimination; ;
  • the implementation of the European AI regulation (AI Act).

At the same time, the Ministry of Justice is preparing the gradual integration of AI tools to assist magistrates and staff, while maintaining that these tools must remain a decision-aid and not replace the judge.  

Confidence level: high. The trends described are well established. On the other hand, French case law remains quantitatively limited and continues to build through the early litigation related to artificial intelligence.