French authors and publishers have filed a lawsuit against Meta for alleged copyright infringement for using their works to train its Artificial Intelligence (AI) model. The National Publishing Syndicate, the Society of People of Letters, and the National Syndicate of Authors and Composers have accused the American tech giant of using protected works without permission.
Vincent Montagne, president of the SNE, stated that they have identified many works from members of the union among the data used by Meta. The lawsuit seeks to protect the publishing sector "from the risks of AI that uses its works and cultural heritage to generate 'fake books' that compete with the originals," explained François Peyrony, president of the SNAC.
This legal action takes place in the context of European regulation, as the AI regulation emphasizes the importance of creators of technological solutions respecting copyright and ensuring transparency in the sources used.
The European Union's Artificial Intelligence Act stipulates that generative AI systems must comply with copyright regulations and disclose the origin of the data used in their development. Furthermore, it classifies AI applications into three risk levels, prohibiting those that pose an unacceptable danger, such as government social assessment systems. It also applies to high-risk applications that must meet specific legal requirements and those that are neither prohibited nor high risk, which largely lack regulation.
