The long-running dispute over ChatGPT’s data records has tilted in OpenAI’s favour. A United States federal court has officially lifted the order that forced the company to indefinitely preserve its chat logs.
The ruling, issued on October 9 by Magistrate Judge Ona T. Wang in the Southern District of New York, ends a contentious chapter in the copyright lawsuit brought by The New York Times against OpenAI and Microsoft.
What Did The Court Decide?
Judge Wang signed an order titled “Stipulation and Order to Terminate OpenAI’s Ongoing Obligations Under the Preservation Order at ECF 33”. The court document confirms that OpenAI’s obligation to “preserve and segregate all output log data that would otherwise be deleted on a going forward basis” ended on 26 September 2025, according to the official court filing in NYT v. OpenAI – Order to Terminate OpenAI’s Preservation Order (October 9, 2025).
This means OpenAI can once again delete most chat logs generated through ChatGPT, except for data already preserved before that date. The court further clarified that previously saved output logs will stay available to The New York Times and other news plaintiffs but that logs linked to users in the European Economic Area, Switzerland or the UK are exempt from this preservation.
The order was the result of a joint motion from both sides. Its a rare moment of agreement in an otherwise tense legal fight that began in late 2023. The initial preservation order had forced OpenAI to store vast amounts of data indefinitely to assist the Times in its copyright claims.
Why Was The Order Controversial?
The preservation order was a controversial topic since it was introduced in May last year. It required OpenAI to retain deleted and temporary chat data from its users, a move the company said infringed on privacy. OpenAI called the measure an “overreach” and warned that keeping sensitive data from millions of users could be risky.
News plaintiffs, led by The New York Times, argued that the order was necessary to prove whether OpenAI used their copyrighted work to train its language models. They claimed ChatGPT sometimes reproduced sections of their articles or attributed false information to their outlets.
OpenAI fought the order for months, and some users even attempted to intervene in court to protect their data, but those motions were denied. Judge Wang ruled that users were not parties to the lawsuit and could not object to the data preservation.
More from News
- Reports Show A 85% Rise In The Use Of Travel eSims
- Is ChatGPT Turning Into An “Everything App”?
- US Government Shutdown Reveals IPO Loophole
- What Does The Recent Discord Data Breach Say About Risks Of Age Verification Laws?
- Why Are Companies Spending More On AI Content When People Don’t Like It?
- Revolut CEO Announces He Is Moving To Dubai: Is The UK Losing Too Many Millionaires?
- World Mental Health Day: How The UK Government Is Investing Into Mental Health
- Experts Share: How The Rise Of Ads Impacts The Quality Of Online Platforms
What Happens To The Saved Chat Data Now?
Under the new ruling, OpenAI must keep previously saved logs intact. The court order explains that “OpenAI will continue to preserve the output log data already preserved and segregated pursuant to the Preservation Order prior to September 26, 2025, except for output log data corresponding to user requests originating from within the European Economic Area, Switzerland, or the United Kingdom.”
It also confirms that OpenAI has identified and preserved data linked to domains that The New York Times and other plaintiffs flagged. These domains include more than 90 websites, such as cryptonews.com.au, deccanherald.com, naijanews.com, liverpooldigest.co.uk, and newsnow.com. The list appears in Appendix A of the court order.
If the plaintiffs identify more domains, OpenAI must “meet and confer in good faith” about adding them. This means The New York Times can still expand its review, but OpenAI is no longer required to blanket-save all ChatGPT logs going forward.
What Does The Outcome Mean For OpenAI And The Times?
In practical terms, OpenAI won this round. The termination of the preservation order means the company can revert to its regular data retention policies. The ruling also removes a technical and financial burden, as indefinite data storage was said to be costly.
For The New York Times, the decision narrows its investigative window. It can continue analysing previously saved data and flagged accounts but cannot demand new logs for unlisted users unless new domains are formally added.
The overall copyright case between the two is still active. The Times and other outlets claim that OpenAI and Microsoft’s AI systems have used their copyrighted articles without consent. Both defendants deny the allegations.
The order makes clear that the latest agreement “does not purport to waive, modify, or otherwise affect the parties’ obligations under Federal Rule of Civil Procedure 37(e).” This means either side could still seek further relief if new disputes arise over deleted data.
The ruling makes this the end of OpenAI’s most restrictive data order so far, but the main copyright case continues. The plaintiffs are still pursuing damages and policy changes to stop AI companies from using protected journalism in their training datasets.
Legal analysts say the court’s latest decision gives OpenAI some breathing space as it continues to face lawsuits over copyright and data use. Meanwhile, The New York Times retains access to the logs it already obtained, keeping its investigation alive even as the “chats case” chapter closes in OpenAI’s favour.
