That must be why literally every single AI-powered code review of completely-human-written work turns up something. >..<

Also, regarding your accusation, you need to contextualize- what language, what use-case, what LLM, etc.?