Newspaper publishers sue Microsoft and OpenAI for copyright infringement

Eight American newspaper publishers have filed a complaint against Microsoft and OpenAI in a New York federal court on Thursday, claiming that the tech companies are reusing their articles without permission in generative artificial intelligence products and falsely attributing inaccurate information to them.

The legal challenge comes four months after The New York Times sued OpenAI for copyright infringement over the ChatGPT chatbot that the startup launched in late 2022. OpenAI said in a January blog post that the case was without foundation, adding that she wanted to support “a healthy society”. news ecosystem. Sam Altman, CEO of OpenAI, said in January that the startup had wanted to pay the New York Times and was surprised to learn of the lawsuit.

In recent months, OpenAI has signed deals with a handful of media companies, including Axel Springer and the Financial Times, allowing the Microsoft-backed startup to draw on publishers’ content to improve AI models. . Googlewhich has its own general-purpose chatbot to answer user queries, said in February that it had entered into an agreement with Reddit this includes the right to train AI models on the platform’s content.

The group of eight newspaper publishers is challenging ChatGPT and Microsoft’s Copilot assistant – available in the Windows operating system, the Bing search engine and other products produced by the software company – for “stealing millions of copyrighted articles from publishers without permission and without payment,” according to the complaint.

Representatives for Microsoft and OpenAI did not immediately respond to requests for comment. The newspaper publishers affected by the lawsuit operate The New York Daily News, The Chicago Tribune, The Orlando Sentinel, The Sun-Sentinel of Florida, The Mercury News of California, The Denver Post, The Orange County Register in California and The Pioneer Press of Minnesota. .

They said OpenAI relied on datasets containing text from their logs to train its large GPT-2 and GPT-3 language models, which can spit out text in response to a few words of human input .

“The current GPT-4 LLM will produce near-verbatim copies of significant portions of publishers’ works when prompted to do so,” the complaint states, showing several examples of ChatGPT and Copilot allegedly doing so.

Publishers said Microsoft copies information from their logs for the Bing search index, which helps inform responses in Copilot. But these results don’t always provide links to newspaper websites, where they can view ads next to articles or pay for subscriptions.

The New York Times case also addressed the issue of OpenAI models regurgitating information from its articles. In its blog post, OpenAI called this behavior “a rare failure of the learning process on which we are continually making progress.”

WATCH: OpenAI CEO Sam Altman: US needs AI policy