OpenAI Must Produce Testimony From Musk Case in Copyright Suits

OpenAI Inc. must produce testimony from the company’s recent trial victory over Elon Musk and X.AI as part of separate, sprawling copyright litigation against the artificial intelligence giant, a federal magistrate judge ruled.

Depositions from OpenAI founders Sam Altman and Greg Brockman, as well as from Microsoft Corp. CEO Satya Nadella and the person designated to testify for OpenAI as an entity, could show inconsistency about the commercial nature of OpenAI’s use of copyrighted works, the US District Court for the Southern District of New York said Thursday.

Magistrate Judge Ona Wang found production relevant to that first factor of the fair use analysis and not burdensome to OpenAI—particularly after the plaintiffs narrowed their initial request.

The order could undermine OpenAI’s bid to downplay the commercial aspect of its use of the material at issue in 19 copyright lawsuits that have been consolidated for pretrial purposes.

Musk’s suit asserted that OpenAI betrayed its mission to benefit the public by morphing into a for-profit business. A jury this week rejected Musk’s claims as brought too late, clearing one of several hurdles for OpenAI to move toward an initial public offering.

OpenAI—along with Microsoft—now faces potentially catastrophic liability if a court deems its use of a vast array of copyrighted works to train AI doesn’t qualify as fair use, a novel question being litigated across the country.

The copyright suits—many of them putative class actions—were filed between 2023 and this year, with discovery in the original 12 cases consolidated in April 2025.

In January, as the Musk case headed for trial, the copyright plaintiffs moved to compel OpenAI to produce Brockman’s notebook in which the company’s president recorded notes about his work. Wang ordered OpenAI to produce portions of the diary in March.

That motion also renewed an earlier request for depositions of six individuals from the Musk case. OpenAI later trimmed the request down to the four depositions, Wang said.

OpenAI “inconsistently” objected to the production by arguing both that the plaintiffs haven’t shown any inconsistencies and that plaintiffs misinterpreted the testimony, Wang said. But only OpenAI has the full transcripts of the depositions, making its misinterpretation argument dependent on the court taking its word over the plaintiffs’, she said.

Wang added that even if the testimony is consistent—"which facially it is not"—OpenAI’s commercialization efforts are still relevant in assessing fair use.

The plaintiffs are represented by Susman Godfrey LLP; Lieff Cabraser Heimann & Bernstein; Cowan, DeBaets, Abrahams & Sheppard LLP; and Boies Schiller Flexner LLP.

OpenAI is represented by Keker, Van Nest & Peters LLP; Williams & Connolly LLP; Latham & Watkins LLP; and Morrison & Foerster LLP.

The case is In Re: OpenAI, Inc. Copyright Infringement Litig., S.D.N.Y., No. 1:25-md-03143, 5/21/26.