Robin Van Swank
Sarah Silverman and two other authors filed class-action lawsuits against OpenAI and Meta, claiming the companies’ artificial intelligence software programs were stealing from their copyrighted works.
The case reflects a growing debate over whether emerging artificial intelligence technology crosses the line to infringe on copyrighted works, which came to a head in a recent congressional hearing featuring OpenAI CEO Sam Altman.
In the lawsuit, Silverman and two other authors, Christopher Golden and Richard Kathrey, claim that OpenAI’s ChatGPT relied on their work for its training dataset.
The authors, who are seeking class-action status, say they did not consent to their works being used in this way, but that they were “absorbed and used to train ChatGPT.”
“In fact, when ChatGPT is triggered, ChatGPT produces a summary of Plaintiffs’ copyrighted works — something that would only be possible if ChatGPT had been trained in Plaintiffs’ copyrighted works,” the lawsuit states.
The lawsuit claims that OpenAI copied Silverman’s book BedwetterGoldens Ararat and of Khatre Sandman Slim. It argues that OpenAI’s language models cannot function without the information in the books and that those models “violate derivative works”. Examples of OpenAI publications include case studies when a user asks to summarize books.
Related: Meta’s threads pass 70M records; Users double in one day
The Meta lawsuit cites the company’s application of their work to LAMA language models and says that “decisions about what textual information to include in the training dataset are deliberate and critical choices.”
The lawsuits were filed in federal court in San Francisco. The plaintiffs seek class action status, as well as statutory and other damages and injunctive relief.
