French publishers and authors recently took Meta to court, accusing the company of using their works without authorization to train its artificial intelligence (AI) models. The three trade associations said they will file a lawsuit against Meta in a Paris court, accusing the company of unauthorized “extensive use of copyrighted works” to train its generative AI models.

Founded in 2004 and headquartered in California, Meta (formerly Facebook) focuses on social media, virtual reality, and AI research and development.

In a joint statement, Vincent Montagne, president of the French National Publishing Union (NPU), which represents book publishers, said Meta's data pools “contain a significant number” of its members' copyrighted works.

The National Union of Authors and Composers, the other party to the lawsuit, which represents 700 authors, playwrights, and composers, emphasized that the lawsuit is intended to protect its members from “the use of their works and cultural heritage to train AIs”. The lawsuit seeks to protect its members from “the use of their works and cultural heritage to train AI”.

The third major party is the French authors' rights group, the Association des Lettres, which filed a claim on behalf of the writers' community, demanding the “complete deletion” of the catalog of data that Meta created without authorization and used to train its AI models.

Meta has not responded to the request. The company has rolled out a generative AI-powered chatbot assistant to users of its Facebook, Instagram and WhatsApp platforms.

The European Union's Artificial Intelligence Act requires generative AI systems to comply with the copyright laws of 27 member states and be transparent about the material they use for training.

This is the latest case of conflict between the creative publishing industry and tech companies around data and copyright disputes.