If the code is obvious crap, it doesn’t matter if AI was involved. If the code is perfectly obviously fine, it doesn’t (directly) matter if AI was involved.

Determining quantity of code is very difficult and expensive. It is often that issues with code don't surface until certain patterns of usage.

This is why judging code quality by judging the author's track record, or the track record of their tools, is very valuable.

If you're vibe coding anything non trivial, your code is gonna be terrible. There's no sugar coating it. There's a wealth of examples on the internet that prove it. It's a waste of time spending any effort evaluating it when we have this evidence, and especially when they didn't spend reasonable effort shipping the code.