- cross-posted to:
- databreaches@lemmy.zip
- cross-posted to:
- databreaches@lemmy.zip
MyLovely.AI, an NSFW AI artwork generation platform, has allegedly been compromised, affecting 106,362 registered users. The 2.1 GB database, containing records from April 2026, was posted by unauthorized individuals on a dark web forum. The incident, characterized as a JSON leak, exposes highly sensitive interactions between users and the AI service, revealing both personal identifiers and explicit generated content.
for the love of god if you’re going to do anything with generative AI (and frankly anything else is a better use of your time) you’d better run that shit locally. maybe don’t give tech companies data about your deepest desires
maybe don’t give tech companies data about your deepest desires
But what if I want my life to be an episode of Black Mirror?
explicit generated content
The allegedly compromised data includes:
Negative prompts (Site filters)
I’m curious, does that mean filters users put in for things they don’t want in their porn, e.g. no redheads? Or are those things the users requested that the platform won’t allow them to generate, e.g. CSAM?
I’m curious, does that mean filters users put in for things they don’t want in their porn, e.g. no redheads?
Likely this. Tho ‘site filters’ may indicate it’s including the default stuff that the site itself is filtering like csam.
Some premium blackmail/extortion material in there probably. I’m no fan of Ai but a cybercriminal who would use sensitive sexual information to blackmail someone would be a real nasty SOB and the victims could be extorted to do bad things.