For years, Meta employees have internally discussed using copyrighted works obtained through legally questionable means to train the company’s AI models, according to court documents unsealed on Thursday. The documents were submitted by plaintiffs in the case Kadrey v. Meta, one of many AI copyright disputes slowly winding through the U.S. court system. The defendant, […] © 2024 TechCrunch. All rights reserved. For personal use only.
Court filings in the case *Kadrey v. Meta* reveal that Meta employees internally discussed using copyrighted works, including books, to train the company's AI models, such as those in the Llama family. Meta has argued that such use constitutes "fair use," while plaintiffs, including authors Sarah Silverman and Ta-Nehisi Coates, disagree. Internal communications show employees considering acquiring books without proper licensing, acknowledging legal risks, and discussing the use of Libgen, a platform known for hosting pirated content. Employees also explored ways to mitigate legal exposure, such as avoiding public acknowledgment of using pirated data. The plaintiffs have amended their complaint, and Meta has retained high-profile litigators to defend the case.
