Comedian Sarah Silverman Sues OpenAI and Meta Over Copyrighted Book Training Data
Sarah Silverman and two authors file class-action lawsuits against OpenAI and Meta, claiming the tech giants illegally use their copyrighted books to train AI language models. The complaints argue that the AI tools generate accurate summaries of the plaintiffs' works, proving unauthorized use.
Comedian Sarah Silverman and authors Christopher Golden and Richard Kadrey sue OpenAI and Meta in a San Francisco federal court, alleging the companies use their copyrighted books to train AI language models without permission. The lawsuits seek class action status on behalf of other writers whose work is allegedly scraped to build tools like ChatGPT and Meta's LLaMA.
The complaints argue that the AI models ingest massive amounts of online data, including copyrighted material taken without consent, credit, or compensation. As evidence, the plaintiffs point out that ChatGPT produces detailed summaries of their specific books, which they claim is only possible if the AI trains directly on their protected texts.
The lawsuit against Meta additionally alleges that the company obtains the authors' books through an online "shadow library" known for hosting pirated materials. These legal challenges highlight a growing backlash from creatives who worry that unchecked AI training practices threaten their livelihoods and intellectual property rights.