OpenAI opposes order to share ChatGPT logs over privacy concerns

OpenAI has asked a federal judge in New York to overturn an order requiring it to provide 20 million anonymized ChatGPT chat logs as part of a copyright lawsuit from The New York Times and other news outlets.

The company said sharing these logs could compromise user privacy, arguing that most conversations are unrelated to the copyright claims.

The lawsuit alleges OpenAI used news articles to train ChatGPT and seeks the logs to determine if the chatbot reproduced copyrighted content.

A magistrate judge previously ordered OpenAI to turn over the chats, stating that privacy would be protected through de-identification measures.

OpenAI has until November 14, 2025 to comply with the order.

🔗 Source: Reuters

🧠 Food for thought

Implications, context, and why it matters.

Details and timing remain unsettled

Magistrate’s order required de-identification, but MDL docket materials cover preserving plus sampling of output logs, and omit parameters for a 20 million‑chat production 1.
OpenAI’s objection triggers district judge review (review by the assigned federal district judge), but the docket (the official case record) lists no stay or Friday deadline still enforceable 1.
This case sits in MDL No. 3143 (a process that consolidates related federal cases for pretrial management) in the Southern District of New York (a federal trial court in Manhattan), where OpenAI was ordered to preserve output log data after concerns about deletion 1.
Until the district judge modifies or stays the order, it is unclear whether 20 million chats will be produced or whether scope will shrink.

AI app firms need litigation-ready logging

Discovery (the evidence-exchange phase of a lawsuit) means conversational AI providers should expect large chat productions, which pushes need for automated redaction (software-based removal of sensitive information) plus secure retention systems 2.
Legal-tech (legal-technology) and privacy vendors can package “litigation-ready AI logging” that blends retention policies with access controls plus audit trails to help AI app builders handle discovery without exposing user data 2.
Judges expect comprehensive data governance based on a preservation order across multiple related cases, which turns compliance infrastructure into a competitive edge for AI startups 1.
Enterprise buyers of conversational AI tools will seek contractual guarantees for discovery-safe architectures, opening new partnership room for security and compliance platform providers.