Title: Sarah Silverman Files Lawsuit Against ChatGPT and Meta for Allegedly Training AI on Her Copyrighted Book
Comedian Sarah Silverman has taken legal action against OpenAI and Meta, along with authors Christopher Golden and Richard Kadrey, accusing them of training their AI models, ChatGPT and LLaMA, on copyrighted material from their respective books. The lawsuits were filed on July 7 in the Northern California federal court.
Silverman, Golden, and Kadrey claim that OpenAI’s ChatGPT and Meta’s LLaMA have been trained using copyrighted content, including excerpts from their published literary works. While OpenAI and Meta have openly stated that they do not train their models on copyrighted material, the authors argue that the training data may have been sourced from Library Genesis, Z-Library, Sci-Hub, and Bibliotik. These internet-based torrent repositories are known for hosting copyrighted books. As evidence, the plaintiffs requested ChatGPT to recite passages from Silverman’s book, The Bedwetter, and it delivered verbatim excerpts from the memoir.
Gizmodo reached out to Silverman, OpenAI, and Meta for comment, but there has been no immediate response. Golden and Kadrey chose not to comment on the matter.
Joseph Saveri and Matthew Butterick, the attorneys representing the authors, have a history of litigating against AI companies. They recently posted on their blog LLM Litigation that they have received numerous concerns from writers, authors, and publishers regarding AI’s ability to generate text similar to copyrighted material, including thousands of books. Saveri and Butterick were also involved in a similar lawsuit filed by authors Paul Tremblay and Mona Awad against OpenAI, making claims similar to those made by Silverman, Golden, and Kadrey.
While the hype surrounding AI has diminished, the technology remains under scrutiny. Although AI models like ChatGPT excel at generating coherent text, the problem lies in the fact that language models can extract pre-existing copyrighted material from various sources and assemble it into an original piece of writing. Lawsuits like this are crucial to defining the boundaries of copyright protection in the age of AI.
Please note: This article has been rephrased to ensure originality while maintaining the essence of the original content.