Abstract
Generative artificial intelligence (GenAI) is now a common feature of many students’ study practices, and universities are grappling with its implications for assurance of learning, assessment, academic integrity, effective teaching, student support, and policy. Yet institutional responses are still often shaped by assumptions about student use, capability, trust, integrity, and learning (e.g., Bearman & Fawns, 2024). This project is part of the overall Student Perspectives on AI in Higher Education Project, a four-university collaboration investigating how higher education students in Australia understand and use GenAI in relation to themselves as learners. It included two large scale surveys in 2024 (Chung et al., 2024) and 2026.
The 2024 phase of the project showed that GenAI had already become embedded in students’ academic practices, with over 80% of surveyed students in four Australian universities having used it for study-related tasks and nearly half using it regularly (Chung et al., 2026). However, the findings also revealed complex patterns of trust, usefulness, self-directed learning, equity, academic integrity, and pragmatic, task-specific engagement. While this 2024 survey provided an initial snapshot of student engagement with GenAI, the rapid pace of technological and institutional change necessitated a follow-up survey in 2026 for an updated understanding of student practices and experiences.
The 2026 survey was adapted from the 2024 instrument and comprised 50 questions and 115 items across demographics, use and usefulness, motivations, assessment and academic integrity, learning and fluency, and perceived impact (Henderson et al., 2026). The survey received 10,222 student responses, with 8,700 students fully completing the survey. The sample included 73% undergraduate students, 75% students aged 18 to 24, 59% women, 29% students who spoke a language other than English at home, 19% first-in-family students, and 24% students who identified as having a disability, health condition, or as neurodivergent.
This Pecha Kucha presents descriptive findings from the 2026 survey and, where appropriate, compares these with findings from 2024. The recent 2026 data suggest that student use of AI for study has become more frequent: 20% of students reported daily use and 35% weekly use, compared with 13% daily and 31% weekly in 2024. The most common uses include editing or improving writing, generating ideas or examples, teaching students how to do something, finding information or conducting research, and getting feedback on work.
Further, the presentation explores new areas of inquiry introduced in 2026, including AI summarisation of course materials, motivations for using or limiting AI, and AI fluency and learning. Among students who used AI for study, the strongest reasons for use were that it made things faster, helped them get unstuck, made things easier, reduced stress, and increased confidence. The strongest reasons for limiting or avoiding AI included concerns about inaccuracy, personal learning values, wanting to do the work themselves, worry about breaking rules, and ethical concerns.
By presenting emerging 2026 findings alongside selected 2024 comparisons, this presentation argues that student voices are essential to developing sustainable, ethical, and educationally meaningful institutional responses to AI in higher education.
Keywords
Generative AI, students, higher education, survey, learning, assessment, academic integrity, AI literacy.
