The ChatGPT copyright case between OpenAI and The New York Times (NYT) has escalated into a high-stakes battle over user privacy and data access. At the heart of the dispute is NYT’s demand for 120 million ChatGPT conversations as part of its lawsuit alleging that the AI tool reproduces Times content without permission.
NYT’s Massive Data Demand
NYT argues that access to 120 million chat logs is necessary to determine how often ChatGPT outputs material that is identical or similar to the publication’s copyrighted articles. According to court filings, the Times believes these logs could prove that OpenAI’s training methods directly incorporate its protected work.
OpenAI Pushes Back with a 20M Offer
OpenAI has countered with an offer to provide 20 million anonymized chats instead of the full 120 million. The company claims the NYT request is overly broad and would create significant privacy risks for millions of users.
The tech firm also highlights the logistical challenge: processing and anonymizing such a vast dataset would require substantial time and resources, potentially delaying the trial by months.
Court Order on Deleted Chats Adds Fuel to the Fire
The dispute intensified in May 2025 when a judge ordered OpenAI to preserve all ChatGPT conversations indefinitely, including those users have deleted.
Previously, OpenAI deleted chats after 30 days unless users opted to save them. The new order means every interaction — past and future — must be stored and potentially made available for legal review.
Privacy Concerns Take Center Stage
OpenAI CEO Sam Altman has compared AI conversations to private talks with a lawyer or doctor, calling for what he describes as “AI privilege” to protect users’ sensitive information.
Altman warns that granting NYT full access to 120 million chats could set a dangerous precedent, eroding public trust in AI tools and opening the door for similar demands in future cases.
Why This Case Matters for the AI Industry
- Data Privacy vs. Copyright Enforcement – The outcome will shape how AI companies handle user-generated data in legal disputes.
- Legal Precedent – A ruling in NYT’s favor could embolden other publishers to seek massive data disclosures from AI companies.
- Operational Impact – If courts force indefinite data retention, AI platforms may need to overhaul storage, compliance, and privacy systems.
Background on the Dispute
The NYT filed its lawsuit in December 2023, accusing OpenAI and its partner Microsoft of copyright infringement. The case is one of the first major legal tests for generative AI, aiming to determine whether training AI on copyrighted material without explicit permission violates intellectual property law.
Possible Outcomes
- Settlement – Both parties could agree on a reduced dataset and confidentiality protections.
- Court Ruling for NYT – Could mandate broader access, forcing AI firms to reveal more of their internal data.
- Court Ruling for OpenAI – Could limit how much user data plaintiffs can demand in copyright cases.
