Sarah Silverman is suing OpenAI and Meta for copyright infringement.

Comedian and writer Sarah Silverman, as well as writers Christopher Golden and Richard Kadrey – are both suing OpenAI and Meta in US District Court, alleging dual claims of copyright infringement.

Among other things, the lawsuits allege that OpenAI’s ChatGPT and META’s LLAMA were trained on illegally obtained datasets containing their functions, taking into account the books Bibliotic, LibraryGenesis, Z-Library. and others from “Shadow Library” websites. “Available in bulk through torrent systems.”

Golden and Kadre each declined to comment on the lawsuit, while Silverman’s team failed to respond in a timely manner.

In the OpenAI lawsuit, when asked to summarize their books and infringe their copyrights, all three provided demonstrations showing ChatGPT. from silverman bed wetter ChatGPT is the first book to summarize the performances, while the Golden Key book Ararat Like Kadre’s book it is also used as an example sandman slim, The claim states that the chatbot “never bothered to reproduce the copyright management information that Plaintiffs included in their published works.”

As for the separate lawsuit against Meta, it is alleged that the authors’ books were accessible in datasets that Meta used to train its LLAMA model, a quartet of open-source AI models developed by the company. Introduced in February.

The complaint explains in steps why the plaintiffs believe the datasets are of illegal origin – in the meta paper describing LLaMA, the company references sources for its training datasets, one of which is called ThePile, which is called EleutherAI. Compiled by the company. The stack highlighted in the complaint was described in the EleutherAI paper as “a copy of the contents of the Bibliographic Private Tracker”. The lawsuit states that Bibliotik and the other “shadow libraries” mentioned are “clearly illegal.”

In both claims, the authors say they “have not authorized the use of their copyrighted books as training materials” for the companies’ AI models. Each of his lawsuits includes six counts of different types of copyright infringement, negligence, unjust enrichment and unfair competition. The authors are seeking legal damages, refund of profits and more.

Attorneys Joseph Savery and Matthew Butterick, representing the three authors, write on their LLMitigation website that they have heard from “concerned writers, authors and publishers.” [ChatGPT’s] Amazing ability to generate text similar to text found in copyrighted text, including thousands of books.

Savery has also prosecuted AI companies on behalf of programmers and artists. Getty Images also filed an AI lawsuit, alleging that Stability AI, which created the AI ​​image generation tool Stable Diffusion, trained its models on “millions of copyrighted images”. Savery & Butterick is also representing authors Mona Awad and Paul Tremblay in a similar case involving the company’s chatbot.

Such lawsuits aren’t just a headache for OpenAI and other AI companies; They challenge the limits of copyright. there’s like we said The Vergecast Whenever someone puts Nilay to work over copyright, we’ll see more lawsuits like this in the years to come.

We’ve reached out to META, OpenAI and the Joseph Savery Law Firm for comment, but they have not yet responded as of the writing of this article.