We ran a history assessment last term which was marked by AI.
We found that essays with more factual errors were getting better grades! What's going on?
It actually has very little to do with AI...