United States of America

NYT, Others Seek Sanctions Against OpenAI Over Copyright

Is AI training on copyrighted content theft from creators or a legally protected fair use?

Published JUL 10

NYT, Others Seek Sanctions Against OpenAI Over Copyright

Story

Above: The New York Times logo displayed on a mobile phone screen with OpenAI logo in the background. Image credit: Idrees Abbas/SOPA Images/LightRocket/Getty Images

The Facts

The New York Times, The New York Daily News and 15 other media organizations filed a federal court motion Thursday seeking sanctions against OpenAI for withholding evidence during the discovery phase of their lawsuit.
The New York Times
Breitbart
Publishers allege OpenAI misled the court for two years by claiming it lacked the ability to search its training datasets and ChatGPT output logs. An April deposition allegedly revealed it had conducted such searches.
Ars Technica
The Tech Buzz
TechCrunch
The publishers also alleged that OpenAI deleted or compressed billions of ChatGPT conversation logs, rendering them unsearchable, in violation of the court's preservation orders.
Exchange4Media
Seeking Alpha
OpenAI denied the allegations, saying it would continue "defending our users' privacy and the long-established principles of fair use." It also characterized the motion as an effort to access private user data as the Times' case weakens.
Variety
The News International
The motion, filed in the U.S. District Court in the Southern District of New York, seeks monetary penalties, attorney fees, and a request that the court treat missing chat log evidence as proof of copyright infringement.
Al Jazeera
The Tech Buzz
The Times, which has spent over $28 million on litigation against AI firms, originally sued OpenAI and Microsoft in late 2023. The case was later joined by outlets like the Daily News, and the Center for Investigative Reporting.
Mlex
Variety
GV Wire

Sources Split

LEFT

CENTER

RIGHT

ESTAB
CRIT

CENTER

PRO
ESTAB

The Spin

Narrative A

AI companies have built billion-dollar products by stealing copyrighted content without paying a dime to the creators who made it valuable. OpenAI even stripped bylines and copyright notices before using publishers' work, and Sam Altman admitted these models can't exist without copyrighted material. When nearly 400 local newspapers have to sue just to get compensated, the system is broken.

Hedgie on X

New York Post

Narrative B

Training AI on copyrighted works is fair use — federal judges in both Bartz v. Anthropic and Kadrey v. Meta ruled exactly that, finding the process highly transformative. The real legal line is piracy and indefinite storage, not the training itself. Publishers demanding licensing fees for AI training are pushing a legal theory courts have already rejected.