By Le Figaro with AFP
Published ,
Update
George RR Martin and John Grisham are among the plaintiffs. Getty Images via AFP / JON KOPALOFF / AFP / LOIC VENANCE
Many writers have taken the creator of ChatGPT to court, accusing him of using their books “without permission” to train his language model.
George RR Martinthe author of the saga Game Of Thronesand other writers launched lawsuits against the California start-up OpenAIwhom they accuse of having used their works to create ChatGPT in violation of their copyright.
In the complaint filed Tuesday in federal court in New York, they accuse the company of using their books “without permission» to train its language model, software capable of producing all kinds of texts on simple request in everyday language. “At the heart of these algorithms is systematic theft on a large scale», ert the lawyers. Among the plaintiffs in this cl action are the Authors Guild (organization that represents authors) and several writers, including George RR Martin and the novelist John Grisham. Many other complaints have been filed by artists, organizations, and coders against OpenAI and its competitors.
Read also“It saves me a lot of time”: these French people who use ChatGPT to make their daily lives easier
Language modelsendanger the ability of fiction writers to earn a living, as they allow anyone to automatically generate for free (or very cheaply) texts for which they would otherwise have to pay authors“, argue the lawyers in Tuesday’s complaint. They also argue that generative AI tools can be used to produce derivative content, which imitates the style of writers. “In an unfair and perverse manner, (…) the deliberate copying (of the work) of the plaintiffs therefore transforms their works into engines of their own destruction», erts the complaint.
Mountains of texts
The guild and authors call for a ban on the use of copyrighted books to train language models “without express permission”, as well as damages. OpenAI did not immediately respond to a request from AFP. The company needed mountains of text found online to train its language model, but never specified exactly which sites and writings were used. Propelled among the AI giants thanks to the success of ChatGPT at the end of last year, OpenAi is already the subject of other similar lawsuits, including an action by a group of computer engineers which also attacks Microsoft, its main investor, and the GitHub platform.
And artists filed a complaint in January against Stability AI, Midjourney and DeviantArt, whose programs were trained from numerous visual works on the internet. In early September, Microsoft announced that it would provide legal protection to customers sued for copyright infringement on content generated with its generative AI tools.