Governing Generative AI in higher education: insights from a National ACPHIS Study

Focus of the Showcase: Presentation of interim research outcomes and an evidence-informed, sector-level perspective on generative AI (GenAI) governance and policy in higher education. Background: GenAI is rapidly transforming higher education, yet institutional responses remain uneven and largely reactive. Universities continue to grapple with academic integrity, data governance, intellectual property, ethical use, and equity concerns (Dempere et al., 2023; Sekli et al., 2024; Ma, 2025). Recent analyses highlight tensions between innovation and compliance, alongside inconsistent policy approaches across institutions (Luo, 2024; Nikolic et al., 2024). Regulatory guidance (e.g., TEQSA, 2025) further underscores the need for coordinated, evidence-based policy development. Description: This showcase presents findings from an ACPHIS-sponsored, cross-institutional project examining GenAI use and governance across 47 universities in Australia and New Zealand. The project analyses institutional policies alongside practices in teaching, learning, and research to identify patterns of convergence, divergence, and institutional readiness. Method: A multi-stage interpretive qualitative design is adopted. Stage one includes two systematic literature reviews and a structured analysis of publicly available university policies. Stage two involves stakeholder interviews and focus groups to validate and refine emerging insights. Evidence: Findings identify three policy types: compliance dominant models with extensive controls; performative educational policies combining pedagogy with surveillance; and a smaller group of sociotechnically mature policies integrating AI meaningfully into teaching and learning. Contribution: The study contributes a rare cross-institutional evidence base and advances socio-technical perspectives on digital transformation. Practically, it informs sector guidance, policy coordination, and capability development for responsible GenAI adoption. Engagement: Participants will respond to a live prompt (e.g., “Where is your institution on the compliance–transformation spectrum?”) using Mentimeter, followed by a short peer discussion to compare institutional approaches.