The community arrived at, drafted and enforce policy forbidding AI-generated content; this usually meant answers because that was the obvious use case for LLMs.

(We were getting absolutely flooded from day 1 with really obviously LLM-generated crap, which represented absolutely zero value add over the OP going to chatgpt dot com and copy-pasting the question there instead. Often this came from users who came back to the site after years of hiatus, who had an established track record in previous answers of being barely able to communicate in English at all; and then they would come to the meta site and insist that they hadn't used an LLM, apparently completely unable to understand how blatantly obvious they were being about it.)

The company repeatedly attempt to force shitty AI-powered features on the community, and also tried to intervene with enforcement of that policy, complaining that false positives would be too damaging. These features got roundly rejected again and again, but this was ignored. Again, much of this represented zero value-add over going to chatgpt dot com directly. Some examples:

https://meta.stackoverflow.com/questions/425162

https://meta.stackoverflow.com/questions/425766

https://meta.stackoverflow.com/questions/425081

https://meta.stackoverflow.com/questions/438910

The intervention against moderator enforcement of policy (an actual moderator action! Ordinary community members can't just delete an answer for being LLM-generated, although they do take almost all the other actions people complain about offsite) got so bad that it led to an actual strike.