From bookshelves to bots: academic integrity, generative AI, and the recurring anxiety of higher education

Scholarly concern about academic integrity has repeatedly intensified during periods of technological change that reshape how knowledge is accessed, reproduced, and demonstrated. Across successive shifts, from print to digital platforms, these changes have generated recurring anxieties about authorship, originality, and academic standards, followed by pedagogical and policy adaptations that normalised these tools. In these moments, academic integrity operates less as a fixed moral standard than as a negotiated boundary between acceptable assistance and unacceptable substitution. Contemporary debates surrounding generative artificial intelligence (AI) echo these patterns while introducing new tensions related to scale, opacity, and cognitive delegation. Large language models (LLMs) challenge assessment systems predicated on visible outputs rather than transparent reasoning processes, exposing misalignments between what higher education claims to value – critical thinking, synthesis, and judgement – and what it measures. Rather than constituting a rupture, generative AI exposes the limits of curriculum, pedagogy, and assessment regimes developed for an era of informational scarcity. This perspective argues that the central issue is not tool use itself, but persistent pedagogical practices that conflate academic integrity with learning. We call for integrity frameworks that prioritise process, reasoning, and epistemic agency in technology-rich academic environments.