It proves nothing. AI is sometimes wrong. So is everyone. Most of the time it's regurgitating publicly-available data.

So rather than assserting it's all wrong, why don't you find one thing in it that is?