Mark Zuckerberg will be deposed as part of a lawsuit from authors including comedian Sarah Silverman accusing Meta of stealing their content to train its artificial intelligence assistant.
US District Judge Thomas Hixson on Tuesday dismissed Meta’s bid to block the deposition of Zuckerberg on evidence submitted by the authors that the tech billionaire is the “principal decision maker” of Meta’s AI tools.
Silverman, along with authors Christopher Golden and Richard Kadrey, filed a class action lawsuit against Meta last year.
The suit claimed that Meta used stolen content from their books via “shadow library” websites to train its AI language model LLAMA without consent or compensation.
The lawsuit is one of many in a growing pile of copyright challenges targeting AI chatbots.
The authors also filed an identical lawsuit against OpenAI and its chatbot ChatGPT last year.
Meta had argued that Zuckerberg did not have unique knowledge of the company’s AI initiatives that could not be obtained from other employees.
The case rests on the issue of fair use, which “will largely turn on the transformative nature of the AI models and the alleged effect on the market for Plaintiff’s books,” Meta’s lawyer wrote in a court filing.
The authors “do not need a litany of depositions, let alone, Mr. Zuckerberg’s deposition, to establish or defend these elements,” the lawyers wrote.
But the court ruled on Tuesday that Zuckerberg is the “policy setter” for Meta’s AI enterprises.
The authors “submitted evidence of his specific involvement in the company’s AI initiatives” and his “direct supervision of Meta’s AI products,” Hixson wrote in his ruling.
The plaintiffs had cited an article in The New York Times from April alleging tech giants like OpenAI, Google and Meta bypassed copyright laws to pull ahead in the AI race.
When OpenAI released ChatGPT in 2022, Zuckerberg found himself behind in the sector – and he “immediately pushed to match and exceed ChatGPT, calling executives and engineers at all hours of the night to push them to develop a rival chatbot,” the article said.
Meta has said the books in its AI training data set came from a publicly available data set, but it did not disclose the exact origins.
Meta and the plaintiffs’ attorney did not immediately respond to requests for comment.
A separate group of authors filed a similar lawsuit against AI startup Anthropic in August.
The class action suit alleged Anthropic used pirated copies of authors’ works to train its chatbot Claude.
