The New York Daily News, The New York Times and other media outlets are asking for “serious sanctions.” against OpenAI, saying the company destroyed evidence and intentionally hid its ability to locate stolen news stories in training data and responses to ChatGPT users, according to a new motion filed Thursday.
Read more Medicare’s AI push snarls patients and doctors in errors and delays
The motion, filed in Manhattan Federal Court as part of an ongoing copyright infringement suit, said Open AI expert John Vincent “Vinnie” Monaco “finally revealed” in an April deposition that the company had engaged in a campaign of deception throughout the lawsuit, including by severely misrepresenting technological capabilities and destroying millions of chat histories.
What precise admissions Monaco made were redacted from Thursday’s filing. But lawyers said the disclosure upended the company’s repeated claims — to both the news organizations and the court — that OpenAI was unable to search its datasets for instances of stolen journalism. OpenAI’s alleged theft and reproduction of journalists’ work is central to the lawsuit, which also names Microsoft as a defendant.
“For two years, OpenAI has been making misrepresentations to the court regarding its ability to search for Daily News content in its training datasets and output logs – key evidence concerning OpenAI’s theft of copyrighted content,” said Steven Lieberman, who represents The News and affiliated papers in Tribune Publishing and MediaNews Group.
“This motion asks the court to punish OpenAI for hiding and destroying evidence showing how ChatGPT was trained on stolen journalism.”
The Daily News and other plaintiffs are seeking unspecified monetary sanctions, special jury instructions and other measures.
The motion, written by Lieberman, alleges that OpenAI “withheld highly relevant evidence, prolonged discovery, inflated expenses, and burdened the Court. And not inadvertently: its corporate representative and at least one of its outside counsel knew.”
The companies’ deception was occurring as OpenAI was arguing the news outlets were invading the privacy of ChatGPT users and publicly claiming that any stolen news stories the chatbot spouted were the result of “a rare bug,” court filings detail. Manhattan Magistrate Judge Ona Wang took seriously the privacy concerns of ChatGPT users, previously ordering multiple layers of protection and ensuring any information exchanged was anonymized.
Read more US home prices hit an all-time high as sales slow and mortgage rates rise
In the lawsuit, the news organizations allege OpenAI is stealing and distorting their copyrighted works, inhibiting their ability to sell their original journalism while also providing ChatGPT users reporting that’s often inaccurate. They have been joined by the Authors Guild, and a litany of best-selling writers are also parties in the complex litigation. OpenAI has admitted a growing number of readers are turning to ChatGPT for news, rather than directly to the journalists who gather it.
Earlier this year, Open AI put forward Monaco as the go-to authority on all matters relating to Project Giraffe, an internal operation purportedly designed to limit large language models, or LLMs, from regurgitating copyrighted works. Judge Wang found in January that he was horribly unprepared for his first deposition, unable to answer basic questions, and ordered him to sit for another in April.
Lawyers for the news organizations said the truth came out when they got another shot.
“This is a case about copying. There is no question that it happened. Nor should there be one about what was copied, how often, or to what end. The evidence is in OpenAI’s training datasets and ChatGPT output logs. But instead of just producing that evidence at the start of the case and focusing on the merits of its fair use defense, OpenAI chose obstruction,” the motion says.
“Only a compelled re-deposition of an ill-prepared witness exposed the facts.”
Lawyers and investigators for the news organizations were required to perform a taxing process of piecing together over 80 million responses from ChatGPT users and other evidence related to Project Giraffe, according to Thursday’s filing.
The news organizations are asking that the jurors who ultimately decide the case be made aware that OpenAI deleted billions of ChatGPT responses, in violation of the judge’s orders to preserve data. The company has signaled in the lawsuit that deleted data cannot be retrieved or restored.
“Serious sanctions are especially appropriate here because OpenAI’s conduct, including the violation of the Court’s preservation Order, was knowing and intentional,” Lieberman wrote in the motion.
Read more Most American Jews don’t feel supported by either party or President Trump, new poll finds