A federal judge in the Southern District of New York rejected OpenAI's attempt to narrow discovery obligations in consolidated copyright litigation, requiring disclosure of 20 million anonymized ChatGPT user logs to news plaintiffs including New York Times and Chicago Tribune. The court found that users' voluntary submission of communications and OpenAI's uncontested ownership of logs meant privacy concerns did not justify limiting discovery to search results rather than the full sample. The ruling exposes OpenAI to broad document production in pretrial proceedings that will shed concrete evidence on what training data the company actually used and how user interactions map to copyrighted materials, with implications for the scope of training-data liability exposure and potential settlement leverage for the news publishers.
271 other tracked developments touching the same claims.