Professors turn to old-fashioned methods to beat AI-generated work for tests

Why don't we test students how well they use AI? To go back to older methods seems inefficient if not worse.