OpenAI Must Turn Over 20 Million ChatGPT Logs, Judge Affirms

OpenAI Inc. will have to turn over 20 million anonymized ChatGPT logs in a consolidated AI copyright case after it failed to convince a federal judge to throw out a magistrate judge’s order the company said insufficiently weighed privacy concerns.

Magistrate Judge Ona T. Wang sufficiently considered privacy concerns against the material’s relevance to the ongoing litigation in her discovery ruling in favor of news organization plaintiffs in five lawsuits, District Judge Sidney H. Stein said in an order Monday. She rejected OpenAI’s arguments it should be allowed to run a search of the 20 million-log sample and produce conversations implicating the plaintiffs’ works, saying no case law requires the court to order the least burdensome discovery possible.

The ruling continues a long, winding road in discovery in the consolidated pretrial proceedings for 16 copyright lawsuits against OpenAI. The multidistrict litigation in the US District Court for the Southern District of New York represents a large clump of the dozens of lawsuits filed against AI companies by content producers of all stripes invoking critical, novel questions about how copyright law applies to AI.

The news plaintiffs, which include New York Times Co. and Chicago Tribune Co. LLC, initially requested a 120 million-log sample in July. OpenAI opposed the motion and offered 20 million—0.5% of its preserved logs—to which the publications agreed. But in October OpenAI informed them it wouldn’t produce the entirety of the de-identified sample and offered the search results instead. Wang sided with the news outlets in November, and denied reconsideration in December.

Stein said OpenAI primarily relied on a securities case that was easily distinguishable from the AI-copyright dispute. There the US Court of Appeals for the Second Circuit blocked discovery of US Securities Exchange Commission call recordings from a defendant in another lawsuit. But that case involved concerns that the original wiretaps were illegal, and the subjects of the calls had stronger privacy interests, as they were recorded surreptitiously, he said. ChatGPT’s legal ownership of the logs, however, is uncontested, and users voluntarily submitted their communications, Stein said.

Susman Godfrey LLP, Rothwell, Figg, Ernst & Manbeck, Loevy & Loevy represent the news plaintiffs. Keker, Van Nest & Peters LLP; Latham & Watkins LLP; and Morrison & Foerster LLP represent OpenAI.

The case is In Re: OpenAI, Inc. Copyright Infringement Litigation, S.D.N.Y., No. 1:25-md-03143, order issued 1/5/26.