News publishers ask court to sanction OpenAI over withheld evidence
A group of newspapers led by The New York Times and the Daily News, joined by seven affiliated titles, filed a motion asking a Manhattan federal court to sanction OpenAI for what they call discovery misconduct. The publishers say OpenAI withheld or concealed the training datasets and ChatGPT logs that would show how their copyrighted news content was used to build the models, and that a recent deposition contradicted the company’s earlier assurances. Daily News counsel Steven Lieberman said OpenAI had spent two years making misrepresentations about whether it searched its training data for the plaintiffs’ work.
The dispute stems from the suit the Times brought against OpenAI and Microsoft in late 2023, one of the highest-stakes tests of whether training generative models on copyrighted material without a license is lawful. Sanctions motions of this kind are procedural rather than a ruling on the merits, but they matter: if the court agrees evidence was improperly handled, it can order the material produced or instruct a jury to draw adverse inferences — both of which would shift leverage in the case.
OpenAI rejected the accusation, saying the claims are “blatantly false” and that the plaintiffs are pressing the fight as their underlying case weakens. The company has previously argued its use of published material is protected fair use.