OpenAI is being sued for allegedly violating copyright and using authors’ books to train ChatGPT

On Wednesday, two American writers filed a lawsuit against OpenAI in federal court in San Francisco, alleging that the business improperly exploited their writings to “train” its well-known ChatGPT generative AI system.

Writers from Massachusetts Paul Tremblay and Mona Awad claimed that ChatGPT violated the authors’ copyrights by mining data that was illegally taken from thousands of books.

An attorney representing the writers, Matthew Butterick, refused to comment. A request for comment was not immediately answered by representatives of the privately funded OpenAI by Microsoft Corp.

Regarding the information used to train cutting-edge AI systems, many legal objections have been made. Source-code owners who are suing OpenAI and Microsoft’s GitHub are among the plaintiffs, as are graphic artists who are suing Stability AI, Midjourney, and DeviantArt.

The defendants in the complaint contend that their systems constitute fair use of works protected by copyright.

Users’ text messages are answered by ChatGPT in a conversational manner. In January, only two months after its inception, it had 100 million active users, making it the consumer application with the fastest growth in history.

Large quantities of data that have been scraped from the internet are used to generate content by ChatGPT and other generative AI systems. Books are a “key ingredient” according to Tremblay and Awad’s complaint because they provide the “best examples of high-quality longform writing.”

According to the lawsuit, over 300,000 books, including those from illicit “shadow libraries” that sell copyrighted books without a license, were included in OpenAI’s training data.

Among Awad’s best-known works are “13 Ways of Looking at a Fat Girl” and “Bunny.” Tremblay’s books include “The Cabin at the End of the World,” which was turned into the M. Night Shyamalan movie “Knock at the Cabin” and released in February.

Tremblay and Awad said that ChatGPT could provide “very accurate” summaries of their books, proving that these works were included in the database.

In the case, a statewide class of copyright owners whose works OpenAI is accused of misusing asks for an undetermined sum in monetary damages.