News organizations led by The New York Times asked a federal judge Thursday to sanction OpenAI for allegedly hiding a 78 million conversation dataset and deleting billions of ChatGPT logs.
Key Points:
- Publishers accuse OpenAI of concealing for over two years its ability to search ChatGPT logs and training data for copyrighted news content.
- The motion seeks to bar a 20 million log sample from trial and cites the deletion of billions of conversations.
- OpenAI denies the allegations and says the publishers want to invade user privacy.
OpenAI Discovery Allegations
The sanctions motion, filed Jul. 9 in Manhattan federal court, also names the New York Daily News, Ziff Davis and the Center for Investigative Reporting as plaintiffs. It accuses OpenAI of spending more than two years misrepresenting its ability to search its own systems for copyrighted news content.
Throughout the case, the company argued it could not search its training corpus. It also said producing ChatGPT conversations would be burdensome and would put user privacy at risk.
That position unraveled in April, when privacy engineer Vincent Monaco revealed under deposition that OpenAI had already run such searches internally. His testimony allegedly showed the firm had amassed roughly 78 million de-identified conversations before the lawsuit began and screened them for news content.
Publishers say none of that was disclosed while they spent months combing a heavily redacted sample of 20 million logs, far below the 120 million they originally requested. The court deemed that sample unusable, and the motion further accuses OpenAI of deleting billions of conversations despite a standing preservation order.
Also Read: Is Charles Hoskinson Leaving Cardano? He Calls The Claim A Complete Lie
Publishers Seek Severe Penalties
The plaintiffs want the court to bar OpenAI from relying on the 20 million log sample at trial. They also ask the judge to find that ChatGPT outputs contained substantial reproduction of their copyrighted work and to block any argument to the contrary. Attorneys' fees and a jury instruction about the deleted logs round out the requested remedies.
Lead counsel Ian Crosby said OpenAI "lied to The Times, The Daily News Plaintiffs, the public, and the court" while concealing searches it had already run. Daily News attorney Steven Lieberman accused the firm of "hiding and destroying evidence showing how ChatGPT was trained on stolen journalism."
Drew Pusateri, a spokesperson for the AI firm, called the claims blatantly false and said the company will "continue defending our users' privacy and the long-established principles of fair use."
The Times sued OpenAI and Microsoft in December 2023, arguing the pair built substitutive products on millions of its articles without permission or payment. The publisher has since spent more than $28 million fighting AI companies in court, a tally that includes a separate suit against Perplexity. Anthropic, meanwhile, agreed last year to pay authors $1.5 billion over pirated books, the largest AI copyright settlement so far.
Read Next: Palantir Sheds 50 Engineers And A Quarter Of Its Value: What Went Wrong?