I'd guess that valid human-authored submissions were being rejected by moderators because they appeared to the moderator to be AI-authored. Moreover, there is probably a significant grey area where a human selects among multiple AI answers and refines them.

As a user of Stack Overflow answers, I don't care too much about who the author was. I do value the voting, comments and community on Stack Overflow, since they add confidence and color to an answer.