Acceptance of generative AI in higher education: A latent profile analysis of policy guidelines

Abstract

Generative AI tools such as ChatGPT and Bard are quickly changing higher education, bringing both opportunities and challenges. This study examines how top-ranked universities differ in their acceptance of generative AI, applying a latent profile analysis to classify universities based on their acceptance levels and four institutional characteristics: the ratio of international students, citation per faculty, academic reputation, and faculty-student ratio. The results revealed four distinct profiles. Profile 1 includes universities with a strong opposition to unauthorized AI use, underscoring academic integrity, and boasting high international student ratios and research output. Profile 2 consists of universities supportive of responsible AI use, despite lower international presence and research output, highlighting the role of a supportive environment. Profile 3 represents universities with a neutral stance on AI, focusing on ethical usage while having strong international presence but struggling with research output. Profile 4 also adopts a neutral stance, with high academic reputations and research output but moderate international presence and lower faculty-student ratios. These findings are in line with previous research on AI acceptance at the student and faculty levels, highlighting the importance of supportive environments and clear institutional policies. This study provides valuable insights for educators, policymakers, and academic institutions navigating the integration of generative AI technologies.