A coalition of news publishers has petitioned a Manhattan federal court to penalise OpenAI, accusing the artificial intelligence company of obstructing discovery in a landmark copyright dispute by concealing and deleting records that could reveal how its systems ingested protected journalism.
In a filing submitted Thursday, the outlets alleged that OpenAI misrepresented its capacity to search its own large language models for copyrighted material, even though such searches had been conducted before litigation began. The plaintiffs further asserted that billions of ChatGPT conversation logs were either erased or rendered inaccessible, and they are requesting financial penalties including reimbursement of legal costs and an adverse judicial finding based on the allegedly suppressed records. Steven Lieberman, counsel for the New York Daily News, characterised the conduct as deliberate concealment of evidence showing ChatGPT was trained on misappropriated reporting.
The legal clash dates to late 2023, when the New York Times sued both OpenAI and Microsoft, and the publisher has since spent over $28 million on the broader fight. The Times recently withdrew a secondary infringement claim from its amended complaint.