On Friday, a coalition of Canadian news agencies sued OpenAI for copyright infringement. The joint lawsuit accuses the company of “capitalizing and profiting” from the unauthorized use of their content for ChatGPT. The lawsuit was filed in the Ontario Superior Court of Justice.
The plaintiffs include CBC/Radio-Canada, Postmedia, Metroland, Toronto Star, Globe and Mail, and The Canadian Press. They demand that OpenAI compensate them for damages, pay any profits that the creator of ChatGPT made from the use of their news articles, and prohibit further use of their content.
“OpenAI routinely violates copyrights and online terms of use by scraping large amounts of content from Canadian media to develop its products, such as ChatGPT,” the media outlet said in a statement (via CBC News). “OpenAI capitalizes on and profits from the use of this content without obtaining permission or compensating content owners.”
In a statement, OpenAI noted its partnerships with news content and its opt-out process, while expressing its belief that its activities fall within the definition of fair use.
“Hundreds of millions of people around the world rely on ChatGPT to improve their daily lives, inspire creativity, and solve complex problems,” wrote an OpenAI spokesperson. “Our models are trained on publicly available data, based on fair use and related international copyright principles that are fair to authors and support innovation. We are working closely with news publishers, including on the display, attribution and links to their content in ChatGPT search, and offer them easy ways to opt out if they wish.”
The new OpenAI search engine is built into ChatGPT. It scans websites and directs users to them for more information. The company said it does not use this data to scan or train its models.
Canadian media outlets have joined a long list of companies, individuals, and other organizations that have sued ChatGPT’s developer for unauthorized training on their work. The list includes (among others) The New York Times, The Intercept, Raw Story, a group of non-fiction authors, and comedian Sarah Silverman.
Earlier this year, OpenAI wrote to the British committee that “it is impossible to teach today’s leading artificial intelligence models without using copyrighted material.” This month, the NYT filed a court declaration as part of its lawsuit, claiming that OpenAI engineers accidentally erased evidence of the company’s AI training data.
OpenAI argues that the use of publicly available online content falls under the fair use doctrine. The Canadian plaintiffs objected to this view, writing that “journalism is in the public interest. OpenAI’s use of other companies’ journalism for its own commercial gain is not. It is illegal.”
