OpenAI Appeals ‘Sweeping, Unprecedented Order’ Requiring It Preserve All ChatGPT Logs

Final month, a federal decide ordered OpenAI to indefinitely keep all of ChatGPT’s knowledge as a part of an ongoing copyright lawsuit. In response, OpenAI has filed an enchantment to overturn the choice stating that the “sweeping, unprecedented order” violates its customers’ privateness.

The New York Instances sued each OpenAI and Microsoft in 2023 claiming that the businesses violated copyrights through the use of its articles to coach their language fashions. Nevertheless, OpenAI said the Instances’ case is “with out benefit” and argued that the coaching falls beneath “truthful use”.

Beforehand, OpenAI solely saved chat logs for customers of ChatGPT Free, Plus, and Professional who didn’t choose out. Nevertheless, in Might, the Instances and different information organizations claimed that OpenAI was partaking in a “substantial, ongoing” destruction of chat logs that would comprise proof of copyright violations. Choose Ona Wang responded by ordering ChatGPT to keep up and segregate all ChatGPT logs that will in any other case be deleted.

In a courtroom enchantment, OpenAI argued that Wang’s order “forestall[s] OpenAI from respecting its customers’ privateness selections.” In keeping with Ars Technica, the corporate additionally claimed that the Instances’ accusations have been “unfounded”, writing, “OpenAI didn’t ‘destroy’ any knowledge, and definitely didn’t delete any knowledge in response to litigation occasions. The order seems to have incorrectly assumed the opposite.”

“The [Times] and different plaintiffs have made a sweeping and pointless demand of their baseless lawsuit in opposition to us,” COO Brad Lightcap said in an announcement. He added that the demand for OpenAI to retain all knowledge “abandons long-standing privateness norms and weakens privateness protections.”

On X, CEO Sam Altman wrote that the “inappropriate request…units a nasty precedent.” He additionally added that the case highlights the necessity for “AI privilege” the place “speaking to an AI ought to be like speaking to a lawyer or a physician.”

now we have been considering just lately concerning the want for one thing like “AI privilege”; this actually accelerates the necessity to have the dialog.

imo speaking to an AI ought to be like speaking to a lawyer or a physician.

i hope society will determine this out quickly.

— Sam Altman (@sama) June 6, 2025

The courtroom order triggered an preliminary wave of panic. Per Ars Technica, OpenAI’s courtroom submitting cited social media posts from LinkedIn and X the place customers expressed issues about their privateness. On LinkedIn, one individual warned their purchasers to be “additional cautious” about what info they shared with ChatGPT. In one other instance, somebody tweeted, “Wang apparently thinks the NY Instances’ boomer copyright issues trump the privateness of EVERY @OPENAI USER – insane!!!”

On one hand, I couldn’t think about having a ChatGPT log delicate sufficient knowledge that I’d care if another person learn it. Nevertheless, folks do use ChatGPT as a therapist, for all times recommendation, and even deal with it as a romantic partner. No matter whether or not I’d personally do the identical, they deserve the correct to maintain that content material personal. 

On the similar time, the Instances’ case isn’t as baseless as OpenAI claims. It’s completely value discussing how synthetic intelligence is skilled. Keep in mind when Clearview AI scraped 30 billion images from Fb to coach its facial recognition? Or stories that the federal government uses images of vulnerable people to check facial recognition software program? Sure, these examples exist outdoors of journalism and copyright regulation. Nevertheless, it highlights the necessity for conversations about whether or not corporations like OpenAI ought to want specific consent to make the most of content material moderately than scraping no matter they need from the web.