Titan Law PLLC

The Magistrate Judge’s Ruling

In the recent decision in the case of The New York Times Company v. Microsoft Corporation, et al. (23-cv-11195), U.S. Magistrate Judge Ona T. Wang denied OpenAI’s motion to compel discovery regarding The New York Times’ use and views on generative AI tools. This ruling provides important insights into the scope of discovery in copyright infringement cases involving AI technologies.

The case centers on allegations that OpenAI unlawfully used the Times’ copyrighted works to train its large language models. OpenAI’s motion sought to compel production of documents related to the Times’ use of third-party AI tools, OpenAI’s products, and its position on generative AI. The company argued this information was relevant to its fair use defense.

Judge Wang rejected OpenAI’s arguments, emphasizing that the fair use analysis focuses on the defendant’s use of copyrighted material, not the plaintiff’s views or practices. The court stated, “The fair use factors are concerned with ‘the copier’s use of an original work,’” citing the Supreme Court’s decision in Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith.

Importantly, the court distinguished between relevant discovery and overly broad requests. Judge Wang explained that while documents concerning licensing discussions might be relevant, “discovery concerning the ‘public benefits [from] the copying’ would be directed to [OpenAI] and the public benefits of its copying, not whether nonparties’ Gen AI tools (which presumably were developed without copying) serve a general public benefit.”

The court concluded by stating, “This case is about whether Defendant trained their LLMs using Plaintiff’s copyrighted material, and whether that use constitutes copyright infringement. It is not a referendum on the benefits of Gen AI, on Plaintiff’s business practices, or about whether any of Plaintiff’s employees use Gen AI at work.” 

Some Practical Thoughts from a Litigator

While commentators have focused on the Times’ expected holier-than-thou protestations about the sanctity of reporter notes and the civilization-ending constitutional implications, let me give an important tip for all litigants who file cases:  litigation is entirely and often excruciatingly unpredictable.  The Times filed a lawsuit and, despite the fact that this particular decision went their way, it cannot complain that the other side wants to make them burn money in defending, arguably, relevant requests after being sued. Also, the next decision may not go their way, and the Times can accept it or cry about it. The latter is going to be much more expensive, that’s all.

There will be more requests, more motions, more legal dollars spent.  While the Times can sit on its high horse and presume the world will bow down to its “paper of record” pedigree, litigation doesn’t work that way. 

Don’t like it?  Don’t file a lawsuit.  

Once you do, you open up everything about your business’ processes, plans, methods, and skeletons for examination and scrutiny.  There’s an old saying about “glass houses” and “stones,” as I recall.

What to Watch For

What the mainstream media will likely skip over because they are lazy generally (using generative AI tools to write most of their posts, in all likelihood) is the concept of “fair use of copyrighted material” and how far that exception extends when it comes to published materials and large language models (i.e., OpenAI’s products like ChatGPT).  

The very same people who read The Times (usually, so that they can tell their friends they read The Times) are essentially committing the same offense the Times is accusing OpenAI of doing – i.e., gathering information from the Times so that they can capitalize on it later.  The former is fine, so long as their credit card is approved by the pay wall.  When the information becomes democratized and integrated by OpenAI in responses that don’t fit the Times’ iron-clawed grip on the “narrative,” the Times appears to turn up their nose in disgust or, maybe worse, get real petty about information – you know, the thing that they didn’t create but now claim ownership over?  The fact that a fire happened on 1^st Avenue is just information that anyone can know.  That the Times wrote a news story about it does not give them ownership of the facts, even if they put it on their letterhead.

We’ll keep following this story.

Titan Law PLLC is here to help you if you find yourself in a “fair use” fight, especially for content creators, podcasters, and alternative journalists. Contact jwilson@titanlawny.com if you or a client have questions involving issues related to these topics.

For more useful info, check out the Blog Archive and Fair Use section.

Some coverage of the story (not from the Times, obviously):

Bloomberg Law – July 18, 2024

Bloomberg Law – August 13, 2024

The case is: The New York Times Company v. Microsoft Corporation et al., U.S. District Court (Southern District of New York)., 1:23-cv-11195.