Fighting the New York Times' invasion of user privacy

Why should OpenAI keep those conversations in the first point? (of course the answer is obvious) If they didn't keep them, they wouldn't have anything to hand over, and they would have protected users' privacy MUCH better. This is just as good as Facebook or Google care about their users' privacy.

They didn't keep temporary chats. They were ordered to keep those as part of this case.

>They didn't keep temporary chats

I thought they did? The warning currently says

>This chat won't appear in history, use or update ChatGPT's memory, or be used to train our models. For safety purposes, we may keep a copy of this chat for up to 30 days.

But AFAIK it was this way before the lawsuit as well.

30 days is perhaps a bit long, but they didn't keep them longer than that. It's pretty clear and reasonable.

The dodgy thing is that they don't now warn users that all chats, including temporary, are now "Bcc: NYT"

The NYT requests samples between Dec 2022 and Dec 2024. The judge order to preserve chats came in effect this summer after OpenAI engineers deleted, claiming mistake, the VM in which NYT layers were processing data.

Dates and the 30 day default retention policy don't add up, when framing things this way.