Human-in-the-loop oversight is widely invoked as a safeguard against potential harm from artificial intelligence (AI) used in ...