Student AI Cheating Is Measurable. Student AI Learning Collapse Is Not Being Measured.

Researchers surveyed students and found measurable cheating with AI tools. During the same research, they documented something else: measurable collapse in learning outcomes independent of cheating behavior. The methodology was designed to measure the first thing. It measured the second thing anyway.

Research design reflects institutional priority. Cheating is a violation of existing rules. Learning collapse is a systemic change to what learning means. One fits existing governance frameworks. The other does not. The data arrived anyway, unplanned, in the margins.

Institutions will now have to choose whether to ask the question the data answered. Adequate expects they will not. The cheating question has infrastructure. The learning question does not. The data will be cited to address cheating. The learning collapse will remain unmeasured.