Mark Zuckerberg to be deposed in Sarah Silverman's AI copyright lawsuit

Mark Zuckerberg is being deposed as part of a lawsuit by writers including comedian Sarah Silverman, who accuse Meta of stealing their content to train his artificial intelligence assistant.

U.S. District Judge Thomas Hixson on Tuesday rejected Meta's attempt to block Zuckerberg's testimony because the authors had presented evidence that the tech billionaire was 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 lawsuit alleged that Meta used stolen content from its books through “shadow library” websites to train its AI language model, LLAMA, without consent or compensation.

The lawsuit is one of many in a growing series of copyright lawsuits against AI chatbots.

The authors also filed an identical lawsuit against OpenAI and its chatbot ChatGPT last year.

Meta had argued that Zuckerberg had no unique knowledge of the company's AI initiatives that couldn't be gained from other employees.

The case involves the issue of fair use, which “will depend largely on the transformative nature of the AI ​​models and the alleged impact on the market for plaintiff's books,” Meta's attorney wrote in a court filing.

The authors “do not need a litany of statements, let alone Mr. Zuckerberg’s testimony, to prove or defend these elements,” the lawyers wrote.

But on Tuesday, the court ruled that Zuckerberg was the “politician” for Meta’s AI company.

The authors presented “evidence of his specific involvement in the company’s AI initiatives” and his “direct oversight of Meta’s AI products,” Hixson wrote in his ruling.

The plaintiffs had cited an April New York Times article alleging that tech giants like OpenAI, Google and Meta had circumvented copyright laws to stay ahead of AI competition.

When OpenAI released ChatGPT in 2022, Zuckerberg found itself behind in the industry — and he “immediately set out to match and surpass ChatGPT by calling executives and engineers around the clock to encourage them to develop a competing chatbot push,” the article says.

Meta stated that the books in its AI training dataset came from a publicly available dataset, but did not disclose the exact origins.

Meta and the plaintiffs' attorney did not immediately respond to requests for comment.

Another group of authors filed a similar lawsuit against AI startup Anthropic in August.

The class action lawsuit alleged that Anthropic used pirated copies of the authors' works to train its chatbot Claude.