Context: AI providers are facing dozens of copyright lawsuits in courts across the world, predominantly filed by book authors and news organizations. Perhaps the most heavily targeted AI platform is OpenAI, which has been sued in the U.S., Canada, Germany and most recently India (January 27, 2025 ai fray article). OpenAI was also the first major target of an AI copyright infringement lawsuit in the EU, when German music rights collecting society GEMA in November alleged it has been unlawfully using song lyrics in the training of and output of its ChatGPT (November 14, 2024 ai fray article). GEMA then sued AI music service provider Suno over similar allegations a couple months later (January 21, 2025 ai fray article). It was also revealed in January that Meta has become a defendant in a lawsuit in the Northern District of California for allegedly knowingly using pirated books to train its AI systems – U.S. District Judge Vince Chhabria last week allowed most of that suit to move forward (March 7, 2025 ruling). But so far (like OpenAI and other AI providers), Meta has not faced much of such litigation outside the U.S.
What’s new: Three trade associations representing authors and publishers, the Société des Gens de Lettres (SGDL), the Syndicat National des Auteurs et des Compositeurs (SNAC), and the Syndicat National de l’Édition (SNE), have sued Meta in the 3rd Chamber of the Tribunal Judiciaire de Paris for the “widespread use” of their members’ copyrighted works to train its generative AI models (March 12, 2025 SGDL press release). Meta’s actions are exploiting “cultural heritage”, the SNAC has alleged.
Direct impact: This lawsuit marks the first major, publicly-known AI copyright dispute to be filed in France. The case could therefore set some significant precedents – for both French and EU AI copyright case law (alongside GEMA’s OpenAI and Suno cases). The complaint is also encouraging other authors and publishers to join in, so we may see other such disputes follow in the near future.
Wider ramifications: While a case of this kind may be new to France, it’s certainly not for Meta. This latest move by the creative industry adds to a growing docket of copyright infringement actions the company is facing. Although it is too soon to say whether this pressure will actually force Meta to adjust its allegedly infringing activities, it will be interesting to see how the French case plays out in parallel with those in the U.S.
SGDL represents authors, SNAC (the National Union of Authors and Composers) represents 700 writers, playwrights and composers, and SNE (the National Publishing Union) represents book publishers.
The plaintiffs’ lawsuit demands that Meta comply with French copyright law, by which they mean the complete removal of unauthorized datasets used to train its AI models.
In a statement yesterday, SGDL President Christophe Hardy said the action they are taking aims to instill a “serious commitment” from AI developers to respect creative works, adhere to the legal framework, and, if necessary, establish fair compensation for the use of copyrighted content. “This is essential to preserve a fragile ecosystem whose richness stems from editorial diversity,” he added.
François Peyrony, President of SNAC, said:
“Through this unprecedented legal action in France, we also seek to pave the way for similar cases to protect authors from the threats posed by AI. These technologies not only exploit their works and cultural heritage for training purposes but also generate ‘fake books’ that compete with genuine works by authors.”
While in a separate statement, SNE President Vincent Montagne said:
“Given the extensive presence of works published by members of the Syndicat National de l’Édition in the datasets used by Meta, we are taking this matter to court to establish the violation of copyright and unfair exploitation. Through this lawsuit, we aim to uphold fundamental legal principles. The development of an AI market cannot come at the expense of the cultural sector.”