Answers are grounded in this paper and cite their sources.
This non-empirical conceptual analysis examines how artificial intelligence (AI) is becoming woven into curriculum production, assessment, feedback, student support, academic integrity, accessibility and institutional decision-making. These systems are often justified through connection: faster feedback, personalised support, scalable services and more efficient learning design. Yet technical connection is not the same as educational connection. AI may leave the visible forms of university learning intact while weakening academic authorship, assessment validity, student trust, staff agency, care, equity and institutional legitimacy. The paper develops refusal-capable design: the distributed institutional capacity to decide when AI should be adopted, constrained, redesigned or refused. It integrates connected learning, responsible innovation, sociotechnical studies of educational technology, care ethics, justice-oriented educational judgment and technology non-use scholarship, using selected public examples illustratively rather than as an empirical sample. Six heuristic refusal zones organise recurring legitimacy tensions, and the Distributed Refusal-Capable University Design Cycle translates those tensions into accountable redesign. The contribution is a practical and theoretical account of refusal as a legitimate design outcome rather than simple opposition to AI.
No takes yet. Share an insight, caveat, or question.
Jason Zagami (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: