The Innovation-Enervation Paradox in higher education: AI and the Dunning-Kruger effect

Artificial intelligence is increasingly framed as the solution to higher education’s persistent challenges, including workload, efficiency, and student engagement. Far less attention, however, has been given to the governance conditions that make such claims possible. This paper introduces the Innovation–Enervation Paradox to analyse how institutional AI adoption can generate epistemic, pedagogical, and organisational strain while being sustained as a marker of progress. Rather than treating AI as a neutral tool, the paper examines how it is authorised, normalised, and defended through distributed governance arrangements across universities. Drawing on Fraser’s account of recognition and redistribution, Butler’s theory of performativity, and Haraway’s concept of situated knowledges, it argues that AI governance is a political process that reorganises authority, expertise, labour, and risk. The Dunning–Kruger effect is then mobilised heuristically to explain how institutional confidence in AI becomes stabilised under conditions of partial and uneven knowledge. Through two heuristic scenarios, prompt engineering as pedagogical innovation and AI-enabled academic integrity as governance, the paper explores how practices framed as innovative can narrow epistemic plurality, redistribute labour, and generate recursive cycles in which strain becomes the rationale for further technological intervention.