The New York Times, the New York Daily News and other U.S. publishers asked a Manhattan federal court on Thursday to impose sanctions on OpenAI in a high-profile copyright lawsuit. The publishers contend the company misled the court by claiming it could not search its systems for copyrighted material, while allegedly having already conducted such searches.
According to the filing, OpenAI previously told the court it was unable to inspect its large language models to filter out copyrighted content. The publishers say the company concealed that it had in fact performed searches and investigations before the first plaintiff filed suit.
The plaintiffs also allege that OpenAI deleted or made inaccessible billions of ChatGPT conversations that were relevant to the case. For that reason they are asking for a range of sanctions, including reimbursement of their attorneys’ fees and a judicial declaration that the company’s chat logs demonstrate unauthorized use of copyrighted works.
Background of the lawsuit
The lawsuit was originally brought by The New York Times in 2023. The paper accused OpenAI and its major financial backer Microsoft of using millions of articles without permission to train the large language models behind ChatGPT. The case is part of a broader wave of litigation: authors, visual artists and record labels have also sued technology companies such as OpenAI, Anthropic and Meta, arguing their intellectual property was used without authorization to train AI systems.
Allegations and legal arguments
Ian Crosby, lead counsel for The New York Times, said in the filing that “OpenAI lied for more than two years to the Times, to the Daily News plaintiffs, to the public, and to the court.” He argued the company had claimed that reviewing ChatGPT’s generated outputs was infeasible, overly burdensome and would implicate user privacy, while failing to disclose that it had in fact performed such searches.
The publishers’ filing states that an OpenAI employee later acknowledged the company had conducted multiple investigations into the plaintiffs’ content. The plaintiffs rely on those assertions when seeking sanctions and court findings regarding the preservation and disclosure of evidence.
Why this matters
The dispute could have significant implications for how companies that build large language models handle copyrighted material and preserve evidence during litigation. If the court finds that OpenAI withheld or destroyed relevant material, it could affect remedies and set a precedent for future cases and regulatory discussions.
Next steps
The sanctions motion is part of ongoing litigation; the court will decide based on the evidence and arguments presented whether sanctions or other remedies are warranted. The development represents a new chapter in the lawsuit that began in 2023 and may influence related litigation against other AI developers.
(Reporting based on Reuters.)
