OpenAI and Microsoft face copyright lawsuits over AI training data

OpenAI and Microsoft are facing legal challenges regarding the methods used to collect data for training artificial intelligence models. In a lawsuit filed in the US District Court for the Southern District of New York, US newspaper publishers allege that the companies engaged in systematic scraping of content, including material protected by paywalls, and bypassed technical access controls.

In a related legal development, authors including George R. R. Martin have filed motions for summary judgment in a class-action lawsuit. The plaintiffs argue that OpenAI's use of copyrighted books does not constitute fair use. The filings allege that OpenAI utilized datasets from LibGen and attempted to obscure these sources by renaming them in research papers.

The authors also claim that OpenAI's technology is being developed to mimic specific writing styles, citing internal discussions regarding the model's ability to complete literary series. Microsoft is named in these proceedings due to its significant financial investment in OpenAI and its alleged ability to oversee the company's operations. OpenAI has filed cross-motions asserting that its training practices are protected under fair use doctrines.