If the judge thinks anonymization works this way, he or she is too incompetent to judge on this case.
Those logs are going to get slurped by another AI, of course.
That’ll be great in the case of my chats where I mainly use ChatGPT for security research and I trigger all kinds of LLM recursion issues and sanitization bypasses.
They can have at it with my logs and they can go fuck themselves at the same time.
The first mistake was thinking OpenAI gives a damn about privacy given they’re more than happy to sell that out to Police. Whether it’s a judge or a hacker, it’s only a matter of time before someone gets a hold of that nuclear bomb worth of data.
If “we removed the names” is sufficient to address privacy issues of the people whose information is involved — which seems pretty questionable — I’d think that the same argument would apply to emails on servers and various forms of chat logs.
There’s gotta be some kind of case law on those.





