The role of artificial intelligence in higher education quality assessment: opportunities and limitations in predicting the quality risk in degree programmes

This article explores the use of artificial intelligence in higher education external quality assessment. Specifically, the study demonstrates the feasibility and accuracy of predicting potential non-full compliance (defined as quality risk) in the quality assessment of higher education degree programmes. Using data from the Catalan University Quality Assurance Agency (Catalonia, Spain) covering the period 2015–2023, the study applies machine learning techniques to identify hidden patterns between 74 quality indicators and the accreditation outcomes of 1.416 degree programme evaluations. This research provides an innovative perspective to optimise and complement institutional assessment processes that are widely established at the international level.