Five major Canadian media outlets sued OpenAI, claiming it illegally captured news to train AI models such as ChatGPT

According to IT Home, five major Canadian news media companies have filed a lawsuit against OpenAI, accusing the company of frequently violating copyright and online terms of use, such as using its news to train ChatGPT. The Canadian Press, the Canadian Broadcasting Corporation, Torstar (owner of the Toronto Star), Postmedia (owner of the National Post), and the Globe and Mail issued a statement saying that OpenAI was crawling a large amount of their copyrighted content to train AI models without asking for permission or providing compensation. In an 84-page statement of claim submitted to the Ontario Superior Court, the five Canadian companies asked OpenAI to compensate for losses and permanently prohibit the company from using its materials without consent. In response, OpenAI said that its model is trained based on public data, based on fair use and relevant international copyright principles, which is fair to creators. "We work closely with news publishers, including displaying, attributing and linking their content in ChatGPT searches, and providing them with the option to exit at any time.