Sarah Silverman, Authors Hit OpenAI, Meta With Copyright Suits

Comedian Sarah Silverman and a group of authors sued OpenAI, Inc. and Meta Platforms, Inc. for copyright infringement, claiming the companies’ artificial intelligence generators used the authors’ copyrighted books as training material without permission.

Silverman—author of “The Bedwetter"—and authors Christopher Golden and Richard Kadrey filed a pair of proposed class actions in the US District Court for the Northern District of California against the technology companies on Friday, alleging copyright infringement, Digital Millennium Copyright Act violations and unfair competition. The authors—seeking a permanent injunction, damages and profits—claimed that when prompted, OpenAI’s Chat GPT generates summaries of their works and many of their books appear in the dataset Meta has admitted using to train their AI generator, LLaMA.

This suit follows a rapid increase in litigation over artificial intelligence. OpenAI and other generative AI companies have faced a number of intellectual property lawsuits in recent months, as well as increased scrutiny from Congress and government regulators. Courts and Congress have yet to decide if using copyrighted material to train generative AI models constitutes infringement, leaving attorneys and copyright holders uncertain if they can stop models from using their works.

“When ChatGPT was prompted to summarize books written by each of the Plaintiffs, it generated very accurate summaries,” said the authors’ complaint against ChatGPT. It said that while the AI generator got some of the details wrong, “the rest of the summaries are accurate, which means that ChatGPT retains knowledge of particular works in the training dataset and is able to output similar textual content.”

The authors also claimed OpenAI and Meta intentionally removed copyright-management information such as copyright notices and titles, creating derivative works of the books.

“Meta knew or had reasonable grounds to know that this removal of CMI would facilitate copyright infringement by concealing the fact that every output from the LLaMA language models is an infringing derivative work,” the authors alleged in their complaint against Meta.

The authors’ complaints also speculated that both AI models were trained on content from websites such as Library Genesis and ZLibrary, also known as “shadow libraries” that make millions of books available for permanent download at little to no cost.

“These shadow libraries have long been of interest to the AI-training community because of the large quantity of copyrighted material they host,” said the authors’ complaint against Meta. “For that reason, these shadow libraries are also flagrantly illegal.”

The authors requested a jury trial in both cases.

The parties weren’t immediately available for comment. Joseph Saveri Law Firm LLP represents Silverman, Golden and Kadrey. Representatives for Meta and OpenAI haven’t been docketed.

The cases are Silverman v. OpenAI Inc., N.D. Cal., No. 3:23-cv-03416, filed 7/7/23 andKadrey v. Meta Platforms, Inc., N.D. Cal., No. 3:23-cv-03417, filed 7/7/23.