The lawsuit is among dozens ongoing in court in the US, as content creators seek greater financial benefits for works used to train AI models that could impact their income.
On June 25, judgeince Chhabria of the San Francisco federal court ruled that Meta's use of millions of books to train artificial intelligence models is in line with the principle of "fair use".
Accordingly, this principle allows the use of part or all copyrighted works for purposes such as study, research, information, journalism, and comments without permission from the author or copyright owner.
At the same time, this is also an important principle in the copyright laws of the US and some other countries, helping to balance the copyright of the author and the right to use works for the benefit of the community.
The ruling could be seen as a victory for technology companies that use copyrighted materials to develop AI.
The lawsuit against Meta was initiated by 13 writers including Ta-Nehisi Coates and Richard Kadrey. The plaintiffs allege that Meta's use of a "dark library" of LibGen, which includes millions of online books, academic papers and comics without permission to train AI models for Llama is a copyright violation.
Meta has argued that the works have been used to develop a transformative technology, in line with the principle of reasonable use.
However, judgeince Chhabria noted that his ruling reflected the authors' failure to make reasonable arguments. He emphasized that the authors have not provided enough evidence to prove that Meta's AI model will reduce the market value of their works (eg, reduce book sales or copyright fees). He emphasized: This ruling does not mean confirming that Metas use of copyrighted documents without prior permission to train large language models is legal.
Judge Chhabria suggested that an argument that could bring victory to the plaintiffs in the Meta lawsuit would be a market thinning, referring to the damage caused to copyright owners because AI products can flood the market with countless images, songs, articles, books...
People can promote AI models to create these works with just a fraction of the time and creativity compared to what is usually required, he said, warning that AI can significantly undermine human motivation to create things in the old way.
This is the second victory in a week for the "big guys" in AI development in the US. Previously, on June 23, William Alsup - a judge in a federal court - also ruled in favor of Anthropic San Francisco startup in a similar lawsuit.
