An investigative journalist who exposed fraud at the Silicon Valley start-up Theranos has sued several leading artificial intelligence companies, alleging they used copyrighted books without permission to train their systems.

John Carreyrou, a New York Times reporter and author of Bad Blood, filed the lawsuit in a California federal court on Monday alongside five other writers. The defendants include Elon Musk’s xAI, Anthropic, Google, OpenAI, Meta Platforms, and Perplexity.

The complaint alleges that the companies pirated books and fed them into large language models that power popular chatbots. It is the first copyright case to name xAI as a defendant.

The lawsuit adds to a growing number of legal challenges brought by authors and other rights holders against tech firms over the use of copyrighted material in AI training.

Unlike several similar cases, the plaintiffs are not seeking class-action status. They argue that class actions favour defendants by allowing them to settle many claims at once for lower amounts. “LLM companies should not be able to extinguish thousands upon thousands of high-value claims so easily,” the lawsuit says.

Anthropic reached the first significant settlement in such a dispute in August, agreeing to pay $1.5bn to a class of authors. The new complaint says individual payouts in that case will be a small fraction of the maximum damages allowed under US copyright law.

Perplexity said it does not index books. Other defendants did not immediately respond.

The case was filed by lawyers at Freedman Normand Friedland, including Kyle Roche, who declined to comment.