OpenAI has been asked to hand over 20 million ChatGPT conversation records as the copyright battle escalates.

2025-12-05 04:40:20

Abstract generation in progress

[Crypto World] There’s a recent court ruling that’s worth paying attention to. Federal Judge Ona T. Wang directly rejected OpenAI’s request, ordering them to hand over around 20 million ChatGPT user conversations (with identifying information removed) to The New York Times.

Why do they have to hand them over? The court’s reasoning is straightforward—these conversation records are key evidence in determining whether ChatGPT copied The New York Times’ copyrighted content, and they are “proportional to the needs of the case.”

Of course, OpenAI is unwilling, mainly out of concern for user privacy. But the judge’s stance is clear: privacy indeed needs to be considered, but when the evidence is clearly relevant and the burden is not excessive, privacy concerns “cannot be the dominant factor.”

Behind this case lies a bigger issue—where exactly are the legal boundaries for AI training data? Especially when it comes to copyrighted content, what rules should govern data usage? OpenAI will likely have a headache over this for a while.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

9 Likes

Reward
9
5
Repost
Share

Comment

0/400

ReverseTrendSister

· 12-05 05:09

Oh wow, OpenAI is really being cornered this time. Privacy vs. copyright—both are backed by big players, but the judge sided with copyright. Just thinking about those 20 million conversations... who knows what's hidden in there? Copyright holders pulled off a beautiful win this round. The worry is that all AI companies might have to get their "pants pulled down" for audits like this. At the end of the day, it's still about the training data—who actually owns it? No one can say for sure. If all that has to be handed over, just imagine how those users must feel... anxious, right? OpenAI really didn’t expect the judge to be this tough—the privacy shield just doesn’t hold up anymore.

View OriginalReply0

FastLeaver

· 12-05 04:50

Haha, OpenAI is really in trouble this time. They can’t play the privacy card. --- 20 million chat records—just talking about that amount of data is wild. The judge is determined to see it. --- To put it bluntly, copyright holders have finally pinned down AI companies. This should have happened long ago. --- How come it’s only now they’ve found a breakthrough on how training data is handled? --- Privacy is privacy, plagiarism is plagiarism. Striking that balance is really tough. --- If OpenAI has to hand it over, just do it. It’s not a big deal if the identity information is removed anyway. --- If this ruling goes through, AI companies will all have to rethink their business in the future.

View OriginalReply0

CodeSmellHunter

· 12-05 04:46

Well, now OpenAI is in a tough spot—they can't play the privacy card anymore. If the court demands data, they have to hand it over. There's nothing more to say. The copyright issue was always going to be settled eventually, and now it's starting. We'll finally be able to see clearly whether ChatGPT really used content from news agencies without permission. AI companies must be getting nervous; there are a bunch of similar lawsuits lining up.

View OriginalReply0

PuzzledScholar

· 12-05 04:40

Privacy has really become a shield now, this move by OpenAI is a bit hard to swallow --- Honestly, 20 million conversations out there, imagine how uncomfortable users must feel --- I have to hand it to the judge, privacy is one thing, but in the face of evidence, you have to give way --- So who would dare to trust ChatGPT in the future... your chat logs might become court evidence someday --- Copyright issues are such a headache, anyway, OpenAI must have absorbed plenty from the cases --- With this verdict out, there will definitely be more lawsuits following, OpenAI has really become the target --- Is it really okay to use data freely once it's de-identified? Sounds easy, but is the cost of reverse engineering really that high? --- Feels like the issue of AI training data will never be sorted out, copyright holders are never satisfied

View OriginalReply0

ruggedNotShrugged

· 12-05 04:40

Ha, OpenAI got pinned down this time—privacy arguments can’t beat copyright claims after all. If you ask me, this should’ve been investigated long ago. Training data has always been a gray area. 20 million data entries—if they really have to hand that over, future lawsuits are going to explode. But honestly, copyright holders can’t hold back anymore either. In the AI era, everyone wants answers. The judge’s ruling here is a direct slap in the face to the “privacy above all” narrative. If this sets a precedent, every AI company will have reason to worry. But wait, will OpenAI really comply so easily? Feels like there’s more to come. Anyway, I think OpenAI had this coming—they quietly consumed so much content...

View OriginalReply0

Trending TopicsView More
#JoinGrowthPointsDrawToWiniPhone17
278.55K Popularity
#DecemberMarketOutlook
75.14K Popularity
#PostonSquaretoEarn$50
11.82K Popularity
#LINKETFToLaunch
12.52K Popularity
#SharingMy100xToken
13.71K Popularity

Hot Gate FunView More

1
MOONMoon
MC:$3.58KHolders:1
0.82%
2
GGPGate Guys Penguin
MC:$3.5KHolders:1
0.00%
3
GDGate Duck
MC:$3.75KHolders:2
0.85%
4
GGPGGP Wallet
MC:$3.6KHolders:1
0.81%
5
谁有实力发一个一起拉谁有实力发一个一起拉
MC:$3.52KHolders:1
0.00%

Sitemap