In a major legal development, a group of 17 prominent authors have filed a copyright lawsuit against OpenAI, the parent company of ChatGPT, alleging large-scale plagiarism.
Among the notable authors involved in the lawsuit are George RR Martin, known for the Game of Thrones series, and John Grisham, the author of A Time To Kill.
The authors allege that OpenAI illegally used their literary works “without permission or consideration” and then used their copyrighted material to train its large language model GPT.
OpenAI CEO Sam Altman has revealed that the development of their latest model, GPT-4, cost an estimated $100 million and was trained on 45 gigabytes of large data.
The focus of the complaint is the allegation of copyright infringement. Initially, ChatGPT provided verbatim excerpts from copyrighted texts, though it later switched to summarizing those texts. However, this change in approach did not alleviate concerns about plagiarism.
The class-action lawsuit seeks a jury trial, as the authors are concerned that AI tools like ChatGPT pose a significant threat to their profession.
This concern is not only widespread in the literary field but also in many different industries, such as the ongoing strikes by writers and actors against major media conglomerates like Hollywood.
These negotiations are aimed at ensuring fair compensation and addressing concerns about AI’s involvement in writing and the use of AI-generated images of actors.
The complaint emphasizes that ChatGPT threatens the livelihoods of writers by allowing anyone to produce content automatically and at low cost, instead of the writers who would be paid to create it.
Furthermore, the authors contend that ChatGPT undermines the market by simulating, summarizing, or interpreting their works, thus constituting copyright infringement.
The list of plaintiffs includes renowned authors such as David Baldacci, Mary Bly, Michael Connelly, Sylvia Day, Jonathan Franzen, Elin Hilderbrand, Christina Baker Kline, Maya Shanbhag Lang, Victor Lavalle, George R.R. Martin, Jodi Picoult, Douglas Preston, Roxana Robinson, George Saunders, Scott Turow, and Rachel Vail.
This lawsuit is not the first legal battle involving OpenAI and ChatGPT. Previously, writer and comedian Sarah Silverman, along with two other writers, filed a copyright infringement lawsuit against OpenAI and Meta, alleging that their works had been used to train ChatGPT and LLaMA.
OpenAI has also faced legal challenges in the past, with a group of anonymous individuals filing a 157-page lawsuit. This lawsuit alleges that the AI company collected approximately 300 billion words from the internet, including books, articles, websites, and posts, without consent.
The plaintiffs argue that this data theft was carried out to enhance AI models for profit.
In addition, The New York Times is reportedly considering filing a lawsuit against ChatGPT for using their archives to train the model. The newspaper is concerned that ChatGPT could be summarizing their content through online access.
In the complaint, the plaintiffs ask for validation of the statements made by Sam Altman and OpenAI before Congress. Altman has previously asserted that OpenAI wants to promote a creative, vibrant economy and does not intend to replace creators.