Meta Platforms, Facebook’s parent company, stopped seeking book licensing deals for AI model training due to a copyright infringement lawsuit filed against them. The case, Kadrey v. Meta Platforms, is part of a series of legal disputes between AI firms and IP owners over copyright concerns.
The court documents include excerpts from Meta employees’ depositions, revealing concerns about the practicality of negotiating AI training data licenses for books. Sy Choudhury, leading Meta’s AI partnerships, noted slow engagement from publishers, despite extensive outreach. Initially, a long list of top publishers was compiled from internet searches. However, many cold call attempts to establish contact did not yield the desired responses.
In April 2023, Meta halted AI book licensing due to timing and logistical concerns noted in court transcripts. Choudhury disclosed that certain publishers lacked rights to the content Meta planned on licensing.
The plaintiffs, like authors Sarah Silverman and Ta-Nehisi Coates, amended their complaint several times since filing in 2023. The recent version claims Meta compared pirated books to licensed ones to assess potential licensing deals. The complaint also accuses Meta of using “shadow libraries” with pirated e-books to train AI models, including the Llama series.
Sources News From Various Digital Platforms, Websites, Journalists, And Agencies.
Leave a Reply