In 2021, amid pandemic school closures, the presenter argued that the “new normal” in education had no settled meaning, and cautioned that technology should serve learning principles rather than be adopted out of fear. This presentation revisits that argument against a second disruption, generative AI. Where COVID-19 tested education at the point of delivery, generative AI tests it at the point of evidence: what student work can still tell us about what a student knows and can do. Drawing on Australian higher education, the presentation traces how lecturers’ confidence in authorship, once high but partly assumed, collapsed after late 2022, and how many institutions turned to AI detection tools that offered only a false recovery. Judged by the standards applied to any assessment, these tools fail tests of validity, reliability and fairness, especially for multilingual students. The presentation argues that generative AI disrupts the forms of assessment rather than its principles, and that disruption works as exposure. It proposes a shift from assurance of authorship to assurance of learning, in which confidence is earned through design and academic judgement, the ground on which any claim to rigour, deep learning or critical thinking must rest. It then names three tensions this shift does not resolve: validity against feasibility, validity against reliability and fairness, and assurance against trust. It concludes that schools of education carry this challenge three times over: as providers of a qualification, as receivers of school credentials, and as the formers of teachers who will make these judgements in classrooms.
- Role of Universities: Dilemmas and Tensions in Navigating Change The future of schooling: strengthening deep learning, academic rigour and critical thinking
- Harold Culala
- Taylor’s University Industry Advisory Panel- School of Education (Online, 15/10/2026)
- 9917581510161
- Office of Learning and Teaching
- English
- Presentation
