Comedian Sarah Silverman sues OpenAI, Meta for copyright infringement

Stand-up comic, writer and actor Sarah Silverman has filed lawsuits against OpenAI and Meta, alleging that these tech giants unlawfully used her copyrighted book to train their artificial intelligence (AI) systems.

The legal action, initiated on July 7, accuses both companies of violating book copyright law by utilising Silverman's and two other plaintiffs' works to enhance their AI systems, namely OpenAI's ChatGPT and Meta's LLaMA. 

Matthew Butterick, one of the lawyers representing the plaintiffs, called for ethical practices in AI.

"AI needs to be fair and ethical for everyone. But Meta is leveraging the work of thousands of authors with no consent, no credit and no compensation."

Silverman's lawsuit against OpenAI highlights that when prompted to summarise her 2010 book, ChatGPT generates a detailed text that would only be possible if it had been trained on the copyrighted works of the plaintiffs. The lawsuit includes the ChatGPT-generated summary of The Bedwetter as evidence, revealing specific details but lacking copyright management information.

The lawsuit against Meta focuses more directly on the training corpus, which refers to the body of text used to teach an AI model communication patterns. Silverman's The Bedwetter is mentioned as part of the dataset Meta acknowledged using in a paper that introduced its large language model, LLaMA.

Represented by Silverman, fantasy writer Christopher Golden, and science-fiction novelist Richard Kadrey, the class-action lawsuits against OpenAI and Meta encompass multiple counts, including copyright infringement, removal of copyright management information, unfair competition, negligence and unjust enrichment. The plaintiffs seek statutory damages and modifications to the AI language models. The lawsuits were filed in the United States District Court in San Francisco, where OpenAI and Meta are headquartered, while Silverman resides in Los Angeles.

The lawsuits are led by Butterick, based in Los Angeles, and the Joseph Saveri Law Firm in San Francisco, known for taking on tech giants in previous legal battles. The firm's track record includes a $52 million settlement with Facebook over content moderators' working conditions and $415 million in settlements with Google, Apple, Adobe and Intel regarding "no-poach" agreements aimed at restraining employee recruitment among competitors.

Now, the law firm aims to make its mark in the evolving legal landscape surrounding intellectual property and AI. Saveri and Butterick are already representing open-source programmers in a lawsuit against OpenAI, Microsoft and GitHub Copilot, alleging violations of the programmers' licenses by the Copilot AI coding tool. They have also filed suits on behalf of artists against AI image generator sites like Stability AI and Midjourney.