Authors Accuse Meta of Using Pirated Books for AI Training

Published On Fri Jan 10 2025
Authors Accuse Meta of Using Pirated Books for AI Training

Meta knew it used pirated books to train AI, authors say | 1330 ...

According to court papers filed by a group of authors, Meta Platforms allegedly used pirated versions of copyrighted books to train its artificial intelligence systems. The authors, including Ta-Nehisi Coates and Sarah Silverman, claimed that Meta had approval from its CEO Mark Zuckerberg for this practice. These allegations were revealed in documents made public on Wednesday in California federal court, as part of a copyright infringement lawsuit against Meta.

Meta's Next Llama AI Models Are Training on a GPU Cluster 'Bigger ...

The authors asserted that internal documents obtained during the discovery process revealed that Meta was aware that the works used for AI training were pirated. The tech giant was sued in 2023 for allegedly misusing the authors' books to train its language model Llama, among other AI products.

Allegations of Misuse of AI Training Dataset

The authors presented new evidence to the court, indicating that Meta utilized the AI training dataset LibGen, which is known to contain millions of pirated works. They claimed that Meta distributed this dataset through peer-to-peer torrents. Internal communications from Meta suggested that despite concerns raised within the company about the pirated nature of the LibGen dataset, Zuckerberg approved its use.

Meta Hit With Class Action for Allegedly Using Pirated Books to ...

In a previous ruling, U.S. District Judge Vince Chhabria dismissed claims of copyright infringement related to text generated by Meta's chatbots and the removal of copyright management information from the authors' books. However, the authors argued that the new evidence supported their infringement claims and warranted revisiting the copyright management information (CMI) claim, as well as adding a new computer fraud claim.

Court's Decision

During a recent hearing, Judge Chhabria expressed willingness to allow the authors to file an amended complaint. While he agreed to revisit certain claims, he also indicated skepticism regarding the fraud and CMI claims put forth by the authors.

ChatGPT and Meta AI are the same? We asked WhatsApp, the answer ...

Meta Platforms has not provided any comments on the allegations raised by the authors in the court filings.

You must be logged in to post a comment. This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.