The Pulitzer Prize-winning investigative reporter who helped bring down Theranos is now taking on Silicon Valley's biggest AI companies, accusing them of looting his books to build multibillion-dollar chatbots.
John Carreyrou, the journalist behind the Wall Street Journal's disgraced blood-testing startup Theranos' exposé, sued six major artificial intelligence companies in California federal court on Monday, accusing them of illegally using copyrighted books to train their AI systems.
Carey Lou, author of the best-selling book “Bad Blood,” along with five other authors filed a lawsuit against Google, Elon Musk's xAI, OpenAI, Metaplatform, Anthropic, and Perplexity.
The complaint accuses the companies of pirating books without permission or compensation and incorporating them into the large language models that power popular chatbots.
The case marks the first copyright case to name xAI as a defendant, and adds to a growing legal attack by authors and publishers over how artificial intelligence systems are trained.
Carreyrou, who now works for the New York Times, and other plaintiffs argue that the AI industry builds its core technology on stolen intellectual property, attracts huge investments and reaps profits, while creators receive nothing.
A Perplexity spokesperson said the company “does not index books.”
The other defendants did not respond to requests for comment, according to Reuters. The Post contacted them.
The case comes amid a wave of copyright lawsuits targeting AI developers who scraped text, images and other works from the internet to train their models.
Unlike other high-profile cases, Careyroo and the other authors intentionally avoid class action lawsuits. Class action lawsuits combine claims and allow companies to negotiate a single settlement.
The authors state that class actions favor defendants by limiting their exposure.
“LLM companies should not be able to simply wipe out thousands of high-value claims at low rates,” the complaint states.
The filing is a direct target of a recent settlement in which Anthropic agreed to pay $1.5 billion to settle a class action lawsuit brought by authors who claim the company pirated millions of books for AI training.
Mr. Carreyrou and his fellow plaintiffs withdrew from the agreement, arguing that it grossly undervalued the rights of authors.
According to the new complaint, Humanity Settlement class participants will receive “a small portion (only 2%) of the copyright statute's statutory cap of $150,000” for each infringed work.
The authors say this result shows why class actions cannot hold AI companies accountable.
Mr. Carreyrou previously condemned Anthropic's actions in court, calling the company's use of pirated books an “original sin” and arguing that the settlement did not go far enough to deter future wrongdoing.
Monday's lawsuit was filed by attorneys at Friedman Normand Friedland, including Kyle Roche. Kyle Roche is the lawyer Careyroux profiled in a 2023 New York Times article.
At a November hearing in the Antropic class action lawsuit, U.S. District Judge William Alsup criticized another law firm co-founded by Roche for organizing the authors to avoid a settlement in favor of what the judge called a “sweeter deal.”
Roche declined to comment on Monday, according to Reuters.
