But in the context of this thread, I would say that using an AI to examine logical inconsistencies is the wrong way to use the tool.

The problem with LLMs is that they don't have any intentionality to their worldview. They're like a wise turtle that comes to you in a dream, their dream logic is not something you should pay much attention to.