Education
Universities redesign assessment for the generative-AI era
Universities are increasingly dividing assessment into AI-enabled work and supervised evaluations that verify students’ unaided abilities.
A review of higher education’s response to generative AI described a shift toward assessment models that separate AI-enabled work from supervised evaluation. Universities including the University of Bath and Cardiff University were reported to be using a “two-lane” approach: some assignments allow or require AI, while closed assessments, examinations or oral defenses test unaided individual performance. The report also said a UK survey found that 94% of 1,054 undergraduates had used generative AI for assessed work, while only 48% felt staff were helping them develop AI skills. The trend matters because institutions are redesigning assessment instead of relying solely on AI detection tools.
Original source: AI Weekly