In a major escalation of the legal war over generative artificial intelligence, renowned investigative journalist and "Bad Blood" author John Carreyrou has filed a sweeping federal lawsuit against the industry’s most powerful players, including Google, OpenAI, and Elon Musk’s xAI. The legal action, filed on Monday in a California federal court, marks a pivotal moment in the battle for intellectual property rights, as it represents the first time Musk’s xAI has been formally named as a defendant in a copyright dispute concerning the training of large language models. Carreyrou, a Pulitzer Prize-winning reporter for the New York Times best known for dismantling the Theranos fraud, is joined by five other prominent writers in accusing the tech giants of "pirating" their copyrighted books to feed the insatiable data requirements of chatbots like ChatGPT, Gemini, and Grok.
The complaint alleges that these multi-billion-dollar companies bypassed legal licensing channels, instead sourcing vast quantities of protected literary works from notorious "shadow libraries" such as LibGen and Z-Library to train their artificial intelligence systems. By doing so, the plaintiffs argue, the defendants have built lucrative commercial products on the foundation of stolen intellectual labor. The lawsuit characterizes the unauthorized use of these texts as a "deliberate act of theft," asserting that the AI models are essentially derived from the creative output of authors who have neither been compensated nor consulted. While tech firms have historically argued that such use falls under "fair use" or "transformative" principles, the Carreyrou filing seeks to dismantle that defense by highlighting the direct commercial competition these chatbots now pose to human creators.
A defining feature of this litigation is the plaintiffs' strategic decision to reject a class-action format, a move that signals a hardening of resolve among high-profile authors. The filing explicitly criticizes recent class-action settlements most notably the $1.5 billion Anthropic deal reached in August 2025 arguing that such arrangements allow tech companies to resolve mass infringement claims at "bargain-basement rates." According to the complaint, authors in the Anthropic settlement received only a tiny fraction, roughly 2%, of the $150,000 statutory maximum per infringed work. By filing individually, Carreyrou and his fellow plaintiffs intend to maximize potential damages and prevent the defendants from securing a "sweatheart deal" that effectively grants them a permanent license to the authors’ backlists for pennies on the dollar.
The legal team representing the authors, led by attorneys at Freedman Normand Friedland, including Kyle Roche, is positioning the case as a fundamental challenge to the "original sin" of the AI industry. During earlier hearings related to the AI copyright landscape, Carreyrou famously described the theft of books to build AI as a foundational transgression that continues to haunt the sector’s ethics. While companies like Perplexity have responded by stating they do not "index books," the broader group of defendants has remained largely silent as the industry braces for a discovery process that could reveal the inner workings of how data is scraped, processed, and monetized in the race for AI supremacy.
This lawsuit arrives at a time when the AI industry is under intense pressure from both regulators and the creative class to find a sustainable model for data acquisition. As the case moves forward in California, it is expected to serve as a bellwether for how individual copyright claims will be weighed against the technological momentum of Silicon Valley. If successful, Carreyrou’s legal challenge could force a massive shift in the AI economy, mandating transparent licensing agreements and fundamentally altering the cost structure for developing the world’s most advanced digital assistants.
Post Comment
Be the first to post comment!
