OpenAI's Deception Alleged
Media Outlets Seek Sanctions Against OpenAI for Alleged Evidence Destruction in Copyright Case
Anonymous
Key Points
1
Court venue discrepancy noted across outlets, with some reports naming Manhattan Federal Court while others cite the Southern District of New York (SDNY) as the court handling the matter.
2
The filings assert OpenAI destroyed millions of chat histories, signaling substantial loss of training-data–related evidence relevant to the alleged copyright violations.
3
A defense-team portrayal in the filings describes a 'campaign of deception' by OpenAI, including misrepresenting its technological capabilities in the lawsuit and during discovery.
4
The submissions contend that OpenAI’s corporate representative and at least one outside counsel were aware of the withholding of highly relevant evidence and still proceeded with the case.
Major news outlets — including The New York Times, The Daily News, The Denver Post, and The Orlando Sentinel — are asking a federal judge to impose serious sanctions on OpenAI, accusing the company of destroying evidence and hiding key technical capabilities in a high-stakes copyright case, according to AJC and Newsday. The filings, submitted in Manhattan Federal Court, allege OpenAI ran a 'campaign of deception' to conceal how it used copyrighted journalism to train its AI systems.
The case also names Microsoft as a defendant. The plaintiffs — which include Tribune Publishing and MediaNews Group properties — say OpenAI's own expert blew the company's cover during an April deposition, according to WTOP. The testimony directly undercut OpenAI's repeated claims that it could not search its datasets for stolen news content.
April Deposition Exposed OpenAI's Hidden Search Capabilities
The turning point came during an April deposition of OpenAI expert John Monaco, according to WTOP. Monaco's testimony revealed that OpenAI could, in fact, search its training data and ChatGPT output logs for instances of stolen journalism. That directly contradicted what the company had told the court throughout discovery — the pretrial process where both sides share evidence.
Lawyers for the news outlets say Monaco's admissions showed OpenAI knew about this capability all along. The filings contend that at least one outside attorney and an OpenAI corporate representative were aware of the withheld evidence and still pushed forward with the company's denials, according to Head Topics.
Millions of Chat Histories Allegedly Destroyed by OpenAI
Beyond hiding its search tools, OpenAI allegedly destroyed millions of chat histories tied to the case, according to AJC. Destroying evidence during active litigation is a serious legal violation. Plaintiffs say this loss directly hurt their ability to prove how often ChatGPT reproduced their copyrighted articles word for word.
The outlets argue the deleted records would have shown the scale of the alleged theft. Without them, they say, OpenAI gets an unfair advantage. The filings describe the destruction as part of a broader pattern of misconduct meant to shield the company from accountability, according to Newsday.
Sanctions Sought Include Special Jury Instructions and Fines
The plaintiffs are asking the judge for three things: monetary sanctions, special jury instructions, and other unspecified relief, according to WTOP. Special jury instructions are significant. They would tell jurors they can assume OpenAI destroyed evidence on purpose — a signal that could heavily influence any trial verdict.
The monetary sanctions amount has not been specified in public filings. But the request signals how seriously the outlets view OpenAI's conduct. The legal fight centers on a core question: Did OpenAI build a billion-dollar product by stealing journalism without paying for it?
A Broad Coalition of Local and National Outlets Push Back
The plaintiffs span some of the biggest names in American news. The New York Times, The Daily News, The Denver Post, The Sun Sentinel, The Orlando Sentinel, and outlets tied to Tribune Publishing and MediaNews Group all joined the filings, according to Head Topics. Their combined reach makes this one of the most sweeping copyright challenges against an AI company to date.
OpenAI has not yet publicly responded to the sanctions motion. Microsoft, still named as a co-defendant, also has not issued a statement. The case is being heard in the Southern District of New York. A ruling on the sanctions request could reshape how AI companies handle evidence in future copyright disputes.
Sentiment
Mostly NegativeNEGATIVENEUTRALPOSITIVE
Publishers
20
Articles
264
Reach
284
A
AnonymousAuthor
