On the 27th of December 2023, The New York Times along with many major authors such as George R.R Martin and John Grisham, sued OpenAI and Microsoft in a multi district litigation in New York, (Case No. 1:23-cv-11106).
Their allegation was that OpenAI had used copyright articles, books and other works without permission, to train the platform’s “Large Language Models” (LLMs), potentially infringing copyrights and violating the Digital Millennium Copyright Act (DMCA).
The Plaintiffs’ Claims:
- That OpenAI deleted millions of copyright works from the web, including pirated sources, such as LibGen, to train ChatGPT.
- That the actual outputs of ChatGPT compete with creators’ markets and harm their revenue, such as summaries or texts that resembled original works.
- That OpenAI allegedly removed copyright management information (CMI) from content, thus obscuring the true ownership and violating the DMCA.
OpenAI’s Defence:
- They claimed “fair use” under the U.S copyright law, arguing that AI training is transformative, rather than a direct reproduction of works.
- That their actions asserted no signifiant market harm, as outputs are generalised and never verbatim copies.
- They denied wilful infringement, though internal communications about deleting LibGen data remain under scrutiny.
A timeline of the case:
March 26th 2025
Judge Stein denies OpenAI’s motion to dismiss The New York Times’ core copyright claims, it was found that factual disputes over copying and market harm required a trial.
April 4th 2025
The US Judicial Panel consolidates 12 cases into the MDL, despite objections from plaintiffs, in order to streamline discovery and facilitate expert testimony on complicated AI technology.
May 2025
The Court issues a sweeping preservation order, requiring OpenAI to retain 20 million ChatGPT user conversations, this raises privacy concerns and OpenAI argues that this violates user privacy commitments.
October 8th 2025
A four hour long hearing takes place on discovery motions, including access to OpenAI’s internal messages and emails about deleting LibGen data.
October 28th 2025
Judge Stein denies OpenAI’s motion to dismiss author’s claims, ruling that ChatGPT’s outputs may be “substantially similar” to copyrighted works, and so the case proceeds.
November 7th 2025
Judge Stein rejects OpenAI’s privacy based objections to producing user conversations that had been de-identified, the judge cited protective order and no privacy risk.
November 24th 2025
Magistrate Judge Ona T.Wang orders OpenAI to disclose all in-house lawyer communications about LibGen data deletion, as they may demonstrate wilful infringement which could potentially trigger damages of up to $150,000 per work.
4 reasons why this case is important.
- Its formation of new historical precedent.
The case is challenging the boundaries of copyright law in the era of AI, it asks if training on copyrighted data is fair use or infringement, a question that is still unresolved.
2. It’s incredibly high stakes.
The case concerns itself with millions of individual works, so the stakes are extremely high from a creative standpoint. From a financial standpoint, the damages could reach billions if wilful infringement is proven, potential causing a larger payout than the $1.5 billion Anthropic settlement in August 2025.
3. Its global context.
The case is taking place all across the globe, mirroring previous international cases involving OpenAI such as their dispute with GEM in November 11th 2025, in which OpenAI was made to pay damages after their use of song lyrics was found to violate copyright regulations.
4. Its cultural debate.
The case underpins a prevalent cultural debate of whether AI’s innovation is having positive or negative implications on quality of life. Those that oppose AI’s development, use the impact on creator’s livelihoods as a medium through which to present their case for AI’s damaging capacity and nature.
Who the case’s outcome will affect the most:
Content Creators
If the plaintiffs win, it could ensure that writers, journalists and artists are fully compensated when AI uses their work, this would protect their income and alleviate concerns over who has the control over their work.
General AI Users
Any imposed restrictions on AI training, caused by a plaintiff win, could limit and reduce ChatGPT’s capabilities. This could lead to raised subscription costs and even a delay in the access granted of new features for students and professionals.
The Tech Industry
A loss in the case could force AI firms to secure more expensive licenses, this could slow innovation or lead to antitrust issues as only the industry giants would be able to afford deals. This means that smaller AI startups could face bankruptcy, and be driven out of business by extortionate litigation costs.
The Media Industry
News outlets’ leverage to negotiate licensing deals could be strengthened, which would help them to secure their subscription and advertising revenue against AI generated summary alternatives. The “hallucinations” that AI is prone to producing would be less frequent, which could improve the general reputation for journalistic credibility in the media.
Conclusively, it’s important to focus on the main questions that this case provokes:
Does AI compromise the ownership of creative goods and IP?
Is AI positively impacting the lives of artists and creators?
If the answer is no, how can it be regulated and improved?
“This isn’t just about copyright, it’s about who owns the future of human knowledge.”
— Prof. Tim Wu, Columbia Law (author of The Curse of Bigness)
https://openai.com/form/copyright-disputes
https://authorsguild.org/news/ai-class-action-lawsuits
https://www.theatlantic.com/technology/2025/11/openai-lawsuit-subpoenas/684861
Table of Contents
Written by George Hocking
- Antitrust
- Artificial Intelligence
- Dispute resolution
- Intellectual Property
- Intellectual Property
- Law in Literature
- Legal News
- Mergers & Acquisitions
- Private Equity
- Uncategorized
© George Hocking, 2025
Leave a Reply