They're banning all AI though, so it looks pretty much anti-AI to me.

I wonder - has it been confirmed that no LLMs for PRs literally means no AI assistance for code?

While I haven't codified it anywhere, the policy I would like is for issues and PR descriptions to have no LLMs - there is no reason to ban code completely though IMO. I would say that would be pro human-communication and a stance I would like a lot.

Good, pro AI people produce poor quality in everything they do. They are the least creative and worst problem solvers. I don't want them near me or my work.