NYT reporter sues Google, xAI, OpenAI over AI training

John Carreyrou, a New York Times reporter and author of “Bad Blood,” has filed a lawsuit against xAI, Anthropic, Google, OpenAI, Meta, and Perplexity, alleging unauthorized use of his and other authors’ books to train AI chatbots.

The lawsuit, filed in a California federal court on December 22, 2025 also includes five other writers as plaintiffs and claims the companies used copyrighted material to train their large language models without permission.

This is the first lawsuit to specifically name xAI as a defendant in an AI copyright dispute.

Unlike other recent copyright cases, the plaintiffs are not seeking class action status, arguing that such lawsuits result in lower settlements for individual authors.

The defendant companies have not responded publicly to the lawsuit.

🔗 Source: Reuters

🧠 Food for thought

Implications, context, and why it matters.

Anthropic settlement is public and preliminarily approved; implications for individual suits

The complaint cites a US$1.5 billion Anthropic deal that is public, with preliminary court approval in place 1.
Terms sit in public filings plus a court-approved site and process 2. Claim (to request payment) and opt-out windows are set 2. Each title earns roughly $3,000 before fees, with a default 50/50 split between author and publisher 1.
Plaintiffs argue that class actions pay less per person. Better outcomes from solo suits remain uncertain and depend on facts.

Rights-clearing intermediaries see growing demand from AI companies

Copyright Clearance Center (a licensing intermediary that helps publishers manage permissions) now supports AI training licenses for participating houses 3. This setup did not exist when early models were trained.
Rights-management software makers that build consent workflows can focus on nonfiction publishers with public AI arrangements 3. Wiley, an academic and professional publisher, disclosed $44 million across two agreements 3. HarperCollins, a trade house, rolled out a nonfiction backlist program (older titles) with authors reportedly paid $2,500 3.
Data brokers with rights-tech firms should team with collecting societies (membership organizations that license works and distribute royalties) or large publishers that publish AI terms. These groups can scale licensing volume fast while legal pressure grows.