OK, sure, but again the claim was:

there is no problem in keeping code quality while using AI

Whether or not human-written code also requires review is outside the context of this discussion, and entirely irrelevant.