[GJim]

> I wonder if the breached data is entering AI corpuses.

One would like to think the creators of AI have been prudent enough to ensure AI output obeys data protection law; however the laissez-faire approach the USA takes to data protection (and the hostility of many Americans on here to the GDPR) suggests otherwise.

[gwerbin]

Wasn't Meta caught using pirate book databases for their training data? No decision maker of importance at any of these companies gives a whiff of a fart about data privacy beyond the bare minimum required by the letter of the law, and only when they think the expected cost of breaking the law would exceed the benefit.