French organizations representing publishers and authors announced on Wednesday that they are initiating legal proceedings against Meta, the parent company of Facebook, following the unauthorized use of their books to train generative AI applications.
The three groups involved—SNE, which represents publishers, and the authors’ and composers’ organizations SGDL and SNAC—issued a statement expressing their concerns over the “extensive use of copyrighted works without permission from their creators and publishers.” They accused the American tech giant of engaging in economic “parasitism.”
Vincent Montagne, the head of SNE, stated, “We have identified numerous works published by SNE members within the dataset utilized by Meta.”
The associations contend that Meta, which operates the social media platforms Facebook, Instagram, and WhatsApp, has unlawfully employed copyrighted material to develop its AI models.
See More: EU hits back with countermeasures against Trump’s ‘unjustified’ steel tariffs
Maia Bensimon, the general delegate of SNAC, remarked, “We are witnessing monumental looting.”
Meta has confirmed that it utilized a database known as Books3, which includes the complete texts of approximately 200,000 books, some of which are in French, to train its Llama large language model.
In a separate legal case in the United States initiated by authors, the company acknowledged using this database until 2023, asserting that the AI training fell under “fair use” of the copyright-protected materials.
French publishers and authors have not publicly disclosed an estimated value of the damages incurred due to Meta’s actions.
Christophe Hardy, head of SGDL, stated that their case in the Paris judicial court “should foster a serious commitment from AI developers to consider the creative industries.” He urged AI developers to “adhere to the legal framework and, where appropriate, seek compensation for the use of works that contribute to” their technology.
About The Author
