Archive — US — techcrunch
Court filings show Meta paused efforts to license books for AI training
Meta has paused licensing discussions with book publishers for AI training data due to slow responses and rights issues, according to new court filings in the Kadrey v. Meta copyright case. Meta staff highlighted difficulties in negotiating these licenses, especially with fiction publishers lacking the rights. The case adds to ongoing disputes between AI companies and copyright holders over the use of copyrighted material for AI models.
Key points
- Meta paused licensing discussions with publishers for AI training data.
- Filings in Kadrey v. Meta provide evidence for the pause, revealing difficulties in engagements.
- Meta's Sy Choudhury noted slow uptake from publishers and rights issues with authors.
- Fiction publishers often could not license their books, complicating negotiations.
- There are multiple ongoing copyright cases against AI firms claiming fair use.
- Plaintiffs allege Meta used 'shadow libraries' of pirated e-books for training models.
Why it matters
- The ongoing debate about AI training data and copyright law reflects broader issues in digital content rights, especially as AI technologies evolve.
- The situation parallels debates around copyright infringement and usage rights observed in other technology sectors, like music streaming.
- In 2025, AI companies may face tightening regulations and scrutiny regarding content sourcing and intellectual property rights.
Meta's pause on book licensing illustrates the challenges in bridging AI development with copyright laws, signaling a significant conflict between innovation and intellectual property rights that will shape the future landscape of AI technology.