New Court Filings Add Credence to Reports of Meta’s "Pausing" Talks with Book Publishers
Background
The case of Kadrey v. Meta Platforms is one of many lawsuits filed against AI companies, pitting these companies against authors and intellectual property holders. The defendants, AI companies, have typically claimed that training on copyrighted content is "fair use," while the plaintiffs, copyright holders, strongly disagree.
Recent Court Filings
On Friday, new filings were submitted to the court, including partial transcripts of Meta employee depositions taken by attorneys for the plaintiffs. These transcripts suggest that certain Meta staff felt negotiating AI training data licenses for books might not be scalable.
Meta’s Outreach to Publishers
According to one transcript, Sy Choudhury, who leads Meta’s AI partnership initiatives, stated that Meta’s outreach to various publishers was met with "very slow uptake in engagement and interest." Choudhury recalled that the company had made a long list of top publishers, but didn’t receive contact or feedback from many of them.
Pausing Licensing Efforts
Choudhury added that Meta paused certain AI-related book licensing efforts in early April 2023 due to "timing" and other logistical setbacks. The company discovered that some publishers, particularly fiction book publishers, did not have the rights to the content they were considering licensing.
Additional Insights
Choudhury noted that Meta has paused licensing efforts on at least one other occasion, when attempting to license 3D worlds from game engine and game manufacturers for its AI research team. The company decided to build its own solution instead.
Amended Complaint
The plaintiffs, including bestselling authors Sarah Silverman and Ta-Nehisi Coates, have amended their complaint several times. The latest amended complaint alleges that Meta cross-referenced certain pirated books with copyrighted books available for license to determine whether to pursue a licensing agreement. The complaint also accuses Meta of using "shadow libraries" containing pirated e-books to train its AI models, including the popular Llama series of "open" models.
Conclusion
The new court filings suggest that Meta may have paused or abandoned efforts to license book content for its AI models due to logistical challenges and concerns over rights ownership. The case highlights the ongoing debate over fair use and the role of AI companies in the copyright landscape.
FAQs
Q: What is the purpose of the Kadrey v. Meta Platforms case?
A: The case is a lawsuit filed by authors and intellectual property holders against AI companies, challenging the use of copyrighted content for AI model training.
Q: What is the main issue at stake in the case?
A: The main issue is whether the use of copyrighted content for AI model training is considered "fair use" or constitutes copyright infringement.
Q: What are the allegations against Meta in the case?
A: The plaintiffs allege that Meta cross-referenced pirated books with copyrighted books, used "shadow libraries" containing pirated e-books, and secured some of these libraries through torrenting, which is a form of copyright infringement.
Q: What is the significance of the recent court filings?
A: The recent filings suggest that Meta may have paused or abandoned efforts to license book content for its AI models, which could have implications for the company’s AI development and training practices.
