Key Points
- U.S. magistrate judge orders OpenAI to provide 20 million ChatGPT logs for a copyright lawsuit.
- The data must be de‑identified to protect user privacy before disclosure.
- OpenAI’s COO Brad Lightcap calls the request an overreach that conflicts with privacy commitments.
- OpenAI has seven days to complete the de‑identification process and comply with the order.
- Deleted conversations and temporary chats are typically removed from OpenAI systems within 30 days.
- The case is the first instance of forced disclosure of AI conversation logs in a copyright dispute.
- Legal experts warn the ruling could set a precedent affecting future AI privacy and discovery practices.
Sam Altman and OpenAI
Background of the Litigation
The New York Times sued OpenAI, alleging that the company used the newspaper’s copyrighted content without permission to train its AI models. As part of the lawsuit, the plaintiffs requested access to a statistically valid monthly sample of ChatGPT output logs covering the period from December 2022 through November 2024. The goal is to examine whether the AI system generated responses that incorporated the Times’ material.
Court Order and Scope of Data
U.S. Magistrate Judge Ona Wang issued a ruling that requires OpenAI to hand over a set of 20 million ChatGPT logs. The order specifies that the data must be stripped of personally identifying information before disclosure, citing multiple layers of protection to mitigate privacy risks. OpenAI was given a seven‑day window to complete the de‑identification process and deliver the logs to the court.
OpenAI’s Response and Privacy Concerns
OpenAI’s chief operating officer, Brad Lightcap, has publicly criticized the request, calling it an overreach that conflicts with the company’s privacy commitments. He emphasized that the demand to retain consumer ChatGPT and API data indefinitely undermines long‑standing privacy norms. While the company acknowledges the court’s authority, it maintains that the order jeopardizes user trust and poses a risk to the confidentiality of past conversations.
Impact on Users
According to the ruling, OpenAI’s standard practice already removes deleted ChatGPT conversations and temporary chats from its systems within 30 days. However, the historical logs covered by the order may contain older user interactions. OpenAI asserts that it has already removed all identifying details from the data set, and the judge’s safeguards are intended to ensure that no personal information can be reconstructed from the released logs.
Broader Implications for AI and Privacy
The case marks the first instance in which a court has compelled an AI developer to disclose user conversation data as part of a copyright dispute. Legal experts note that the decision could set a precedent for future discovery requests involving AI‑generated content, potentially reshaping the balance between intellectual property enforcement and user privacy. The outcome may influence how AI companies design data retention policies and privacy protections moving forward.
Next Steps
OpenAI has indicated that it will continue to appeal the order while working to meet the de‑identification deadline. The litigation remains ongoing, and the final resolution will determine whether the disclosed logs can be used to substantiate the plaintiffs’ claims or if additional legal safeguards will be required.
Source: techradar.com