I think an issue with the discourse about #OpenSource software accepting or not accepting #AI code is that it's not hard to get code that's passable as human-written - more so if you consider the quality of human-written code that projects sometimes received before #LLM s got reasonably effective. Whether a project accepts AI code or not, it's near inevitable for AI-written code to be submitted and merged.
What do we do about that? I don't know. It feels like projects would need to start requiring license agreements for patches that effectively absolve them of any infringement claims and put it on the (potentially anonymous, otherwise) submitter or something.
I have no idea what I'm going to do with my own small projects (which have no PRs). It's something I've been thinking about though. I hope other projects are too.