A coalition of French publishers and authors’ societies has filed a lawsuit against Meta, accusing the tech conglomerate of large-scale copyright infringement in the training of its artificial intelligence (AI) models.
The plaintiffs allege that Meta, the parent company of Facebook, Instagram, and WhatsApp, has unlawfully utilised copyrighted works without seeking permission, thus violating established intellectual property rights.
Legal Action and Allegations
The legal action, initiated in a Paris court, unites three prominent French organisations: the National Publishing Union (SNE), the National Union of Authors and Composers (SNAC), and the Society of Men of Letters (SGDL). These groups assert that Meta has engaged in massive breaches of copyright by using their members’ books and other copyrighted materials to train its generative AI model, Llama.
Vincent Montagne, president of the SNE, has publicly criticised Meta for “noncompliance with copyright and parasitism.” Maia Bensimon, the general delegate of SNAC, characterised Meta’s actions as “a significant act of theft.” Echoing this sentiment, Renaud Lefebvre, director general of SNE, described the lawsuit as “a bit of a David versus Goliath scenario,” as they collectively take on a tech giant.
This lawsuit is not an isolated incident but rather part of a growing global debate surrounding the use of copyrighted material in AI training. Similar legal challenges have emerged in the United States, where authors such as Sarah Silverman and Christopher Farnsworth have filed lawsuits against Meta for allegedly using their literary works to train Llama.
OpenAI, the creator of ChatGPT, also faces multiple copyright lawsuits across the United States, Canada, and India.
The lawsuits raise fundamental questions about the scope of “fair use” in the context of AI training and the need for transparency and licensing agreements.
Industry Response
Tech companies, including Meta, have generally argued that their use of copyrighted material falls under the umbrella of “fair use.” They claim that using copyrighted works for AI training is transformative and does not infringe on the rights of copyright holders. However, content creators and publishers dispute this claim, arguing that the unauthorised use of their works undermines their ability to control and monetise their intellectual property.
The European Union is taking a proactive approach to address these concerns. Under the EU’s Artificial Intelligence Act, generative AI systems must comply with the bloc’s copyright laws and be transparent about the material used for training. The French publishers’ lawsuit against Meta also alleges violations of these EU rules.