OpenAI has asked a New York federal judge to deny authors’ demands for disclosure of its recently announced $1 billion licensing agreement with Disney, arguing that the deal has no connection to the copyright claims at the center of the ongoing litigation.
In a letter submitted Thursday to U.S. Magistrate Judge Ona T. Wang, OpenAI said requests for documents tied to the Disney agreement fall well outside the scope of the consolidated copyright cases brought by authors who accuse the company of unlawfully using their written works to train artificial intelligence models.
According to OpenAI, the Disney arrangement does not involve text-based materials and therefore has no relevance to claims concerning books and other written content.
OpenAI told the court that the agreement, which remains subject to final approvals and negotiations, covers only the use of certain Disney-owned characters in video and image-generation tools. None of the AI models challenged by the authors rely on that content, the company said.
“Even if a final agreement exists, it would not bear on the issues raised in this case,” OpenAI argued, adding that the discovery request violates previously negotiated limits on what information must be exchanged during the litigation.
The company accused the authors of attempting to revisit discovery boundaries that were carefully agreed upon earlier in the case.
OpenAI also said producing drafts, communications, and licensing terms related to the Disney deal would impose an undue burden so close to the discovery deadline. Any such materials would require extensive review and redaction of privileged information, the company said, outweighing any potential relevance.
While the authors argue the Disney agreement could shed light on the commercial nature of OpenAI’s business and the feasibility of licensing written works, OpenAI countered that it has already produced financial records addressing commercial use.
“A licensing agreement unrelated to plaintiffs’ works and unrelated AI models does not inform the fair-use analysis here,” OpenAI said.
The authors have also suggested the Disney deal supports claims that OpenAI knowingly infringed copyrights by showing the company understood the need for permission. OpenAI rejected that logic, saying willfulness must be evaluated based on what the company knew at the time the AI models were trained, not on later agreements involving different technologies and content types.
OpenAI emphasized that the alleged harm in the case concerns the market for written works, not animated characters or visual outputs.
The dispute is part of a broader multidistrict litigation pending in New York federal court, where authors allege OpenAI and its partner Microsoft improperly used copyrighted books and articles to train large language models.
The authors seeking discovery include prominent novelists and nonfiction writers, who argue that licensing agreements with major media companies could affect the value and licensing market for their works.
The court has not yet ruled on the discovery request.
Case: In re: OpenAI Inc. Copyright Infringement Litigation
Case No. 1:25-md-03143
U.S. District Court for the Southern District of New York
