Nearly 300 authors file individual copyright claims against Meta over pirated books and its Project Vellum scanning programme

What happened

On October 19, 2026, nearly 300 authors filed individual copyright-infringement claims against Meta Platforms in the U.S. District Court for the Northern District of California, alleging Meta trained its AI models on pirated copies of their books and, more recently, on physical books it bought and scanned without a licence. The claims are spread across five complaints, each named for its first-listed plaintiff: Winspear, Levine, Reno, Sheard and Older v. Meta Platforms. All five make substantially the same allegations.

The individual-claims strategy

Rather than join a class action, each author is pursuing an individual claim. The authors signed up through ClaimsHero, a platform that partnered with the plaintiffs' firms to recruit authors to opt out of the $1.5 billion Anthropic author settlement and sue on their own. The plaintiffs' firms, Freedman Normand Friedland and Stris & Maher, say they represent more than 2,200 authors and plan to file up to 1,000 individual claims by the end of the year. The plaintiffs seek statutory damages of up to $150,000 per work for wilful infringement, to be set by juries, compared with roughly $3,000 per work under the Anthropic settlement.

What the complaints allege

• Meta downloaded books from shadow libraries including LibGen, Z-Library and Anna's Archive, and from a pirated dataset known as Books3, often using BitTorrent, and stripped copyright notices from the files.
• Starting around mid-2025, after it had already been sued over its Llama models, Meta ran a book-scanning programme it called Project Vellum: it bought physical books in bulk, scanned them, and used the text to train its Muse Spark and Muse Glimmer models, released in April and August 2026 respectively. The complaints argue that buying a book conveys ownership of that one copy, not the right to reproduce it.
• Market dilution is expected to be central: the argument that a flood of AI-generated books undercuts sales of human-written ones. Public court filings are scheduled for December 16, 2026.

Status

These are allegations in newly filed complaints; no court has ruled on them. Meta largely prevailed in the earlier Kadrey v. Meta class action, where Judge Vince Chhabria found training to be fair use but indicated he might have sided with the authors had they shown market harm. Judge Chhabria now oversees all the related Meta cases, with related jury trials scheduled to begin May 3, 2027.