WME backs New York Times lawsuit against OpenAI, arguing AI training needs licenses
The talent agency points to an existing licensing market as entertainment groups press for permission and compensation. The court has yet to resolve the dispute.
Loading page…
The talent agency points to an existing licensing market as entertainment groups press for permission and compensation. The court has yet to resolve the dispute.
Listen to this story
WME is urging the court to require licenses and compensation for copyrighted material used in AI training, pointing to OpenAI’s deals with News Corp and Disney as proof that licensing creative content is possible. The brief adds entertainment-industry pressure to The New York Times’ case, whose core copyright claims survived an April 2025 ruling but remain unresolved. The dispute could shape both training-data rights and responsibility for AI outputs that reproduce protected work; reply briefs are due Friday before the judge decides whether to rule or proceed to trial.
WME says OpenAI’s licensing agreements with News Corp and Disney undercut its fair-use defense, while stopping short of calling for a blanket ban on AI training.
The Recording Industry Association of America, music publishers, performance-rights groups, and unions have also filed briefs supporting the Times.
The unions say the copyright dispute affects the rights and livelihoods of more than 250,000 entertainment and media workers.
Talent agency WME has filed a brief backing The New York Times in its copyright lawsuit against OpenAI and Microsoft, adding its voice to entertainment groups demanding payment for AI training. Its argument centers on a concrete challenge to OpenAI’s defense: a market for licensing creative work already exists.
WME submitted its amicus brief on Friday night, Page Six reported on November 3. The filing follows submissions by talent agency CAA and the unions SAG-AFTRA and the Writers Guild of America. These outside interventions bring Hollywood’s concerns about copyrighted work into a case that began with the use of newspaper journalism.
WME disputes OpenAI’s position that using material to train its systems qualifies as fair use. The agency points to OpenAI’s licensing agreements with News Corp and Disney as evidence that content and intellectual property can be licensed. Those arrangements form part of WME’s argument against allowing the company to use protected work without securing rights.
The agency is not calling for a blanket ban on AI companies using creative works. It says companies should obtain licenses and pay fair compensation, according to JKN.
That demand runs against the AI industry’s practical objection to licensing at internet scale. OpenAI and other AI companies argue that checking ownership and securing individual licenses across vast amounts of online content is too complex. JKN reports that the Justice Department has also supported the industry’s argument about the difficulty of individual licensing.
The unions frame the issue as protection for the rights and livelihoods of more than 250,000 entertainment and media workers.
The Times filed its lawsuit on December 27, 2023, alleging that OpenAI and Microsoft used millions of its articles without permission for AI training and services. In April 2025, the court allowed core copyright claims to continue while dismissing some claims concerning copyright-management information and unfair competition. That decision kept the central dispute alive; it did not resolve it.
The newspaper sought summary judgment in September, asking for a decision without a trial. Page Six reports that reply briefs are due Friday, after which the judge could rule or send the case to trial. The pending issues reach beyond the training material itself: they also concern legal responsibility for AI outputs that reproduce protected work.
Loading discussion...
Join the conversation
Explain where you would draw the line between permission and fair use.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.