A group of Canadian news and media companies filed suit against OpenAI on Friday. They claim that ChatGPT developers have infringed their copyrights and unfairly enriched themselves at their expense. The companies behind the suit, which includes the Toronto Star, the Canadian Broadcasting Corporation and the Globe and Mail, are seeking damages and a ban on future use of OpenAI’s copyrighted material.
The news companies said OpenAI used content harvested from their websites to train the massive language models that power ChatGPT. That content is “the product of significant time, effort and expense by the news organizations and their journalists, editors and employees.”
In their lawsuit, the companies wrote: “Instead of trying to obtain the information lawfully, OpenAI has chosen to brazenly misappropriate the valuable intellectual property of news media companies and use it for its own purposes, including commercial purposes, without consent or compensation,” Silverman said. OpenAI also faces copyright lawsuits from authors such as The New York Times, The New York Daily News, YouTube developers and comedian Sarah.
While OpenAI has licensing agreements with publishers such as The Associated Press, Axel Springer and Le Monde, the companies behind the new lawsuit said they have “never received any form of consideration, including payment, from OpenAI for OpenAI’s use of their work.”
An OpenAI spokesperson said in a statement that ChatGPT is “used by hundreds of millions of people around the world to improve their daily lives.” To “stimulate creativity and solve difficult problems,” the company said its models are “trained on publicly available data and based on fair use and related international copyright principles that are fair to developers and support innovation.” Among other things, we provide publishers with an easy way to display, map and link their content in ChatGPT search and to opt out if they wish,” the spokesperson said.
The new lawsuit comes shortly after Columbia University’s Tor Center for Digital Journalism released findings that found “no publisher, regardless of its level of ties to OpenAI, was immune from inaccurate representation of its content on ChatGPT.”
