Two prominent US authors, Sarah Silverman, a comedian and writer, and Christopher Golden, along with Richard Kadrey, have filed lawsuits against OpenAI and Meta, owned by Sam Altman and Mark Zuckerberg respectively. The authors claim that both OpenAI’s ChatGPT and Meta’s LLaMA models were trained on datasets that contained their copyrighted works without permission. The authors allege that their works were obtained from shadow library websites, such as Bibliotik, Library Genesis, and Z-Library, which distribute books en masse through torrent systems.
The lawsuits state that when prompted, ChatGPT generates summaries of the copyrighted works, implying that the model was trained on their books. The authors further argue that the chatbot fails to reproduce any copyright management information included in their works. In a separate lawsuit against Meta, it is claimed that the authors’ books were accessible in the datasets used to train LLaMA. It has been stated that Meta admitted to utilizing these datasets.
Sarah Silverman holds a registered copyright for her book The Bedwetter, while Christopher Golden possesses registered copyrights for several books, including Ararat. Richard Kadrey also owns registered copyrights for various books, including Sandman Slim. In both lawsuits, the authors maintain that they did not provide consent for their works to be used as training materials for the AI models developed by OpenAI and Meta.
The lawsuits consist of six counts, including copyright violations, negligence, unjust enrichment, and unfair competition. The authors seek statutory damages, restitution of profits, and other forms of compensation. Neither Meta nor OpenAI has commented on the lawsuits thus far.
The allegations made in the lawsuits suggest that OpenAI and Meta used datasets containing the authors’ copyrighted works without authorization. These datasets were purportedly sourced from shadow libraries that distribute books through torrent systems. The authors claim that their works were utilized in training the ChatGPT and LLaMA models, leading to summaries being generated by ChatGPT. The lawsuits emphasize the absence of copyright management information within the AI-generated content. Additionally, the authors allege that Meta incorporated their books into the datasets used for training LLaMA.
Sarah Silverman, Christopher Golden, and Richard Kadrey, along with their legal representation, have taken legal action against OpenAI and Meta, asserting copyright infringement, negligence, and unfair competition. The authors seek remedies, including damages and the restitution of profits. OpenAI and Meta have yet to provide any official statements regarding the lawsuits.
