nostr:npub1te88hqz88e6dwu5tskrptdazrvhvf6xdysyawye5j9xnewuwqfksn8u0yq

More to the point, IMO, is that there is nothing legally stopping them from lying and in fact using user input to train their LLM's. How would we know? And what could we do about it after the fact anyway? Without strict legal prohibitions on data collection, combined with guillotine-level enforcement, they are under no obligation to not lie, if lying suits them.

It's like they could replace the whole TOS with "We'll do what we want. Suck it up, loser."

None of these companies should be trusted, ever.