MAPPING GLOBAL GENERATIVE ARTIFICIAL INTELLIGENCE GUIDELINES IN HIGHER EDUCATION: THE AMBIGUOUS BALANCE BETWEEN INNOVATION AND REGULATION
C. Dell'Erba1, L. Ruffini1, R. Silva2, L. Consoli3
Artificial Intelligence (AI), especially Generative AI (GenAI), is now integral to academic life for all stakeholders. Students use it daily to support and optimize their learning. Teachers are mainly realizing its potential for personalized learning and automated assessment. However, some drawbacks need to be highlighted. (Gen)AI is often used uncritically and unethically, leading to critical issues related to academic integrity and social and environmental sustainability. This presents an important challenge for education and shows the strong need for (Gen)AI literacy and policies.
As pivotal centres of innovation and education, universities have a unique responsibility to guide (Gen)AI's ethical and effective use. Because of this, universities worldwide are working to provide adequate guidelines. The current landscape of institutional (Gen)AI guidelines is becoming more explored. However, it is still fragmented and incomplete, creating an urgent need for clarity and systematic analysis. This research investigates how the world's leading universities are formulating and communicating (Gen)AI-related guidelines, specifically focusing on their actions to address the complex (Gen)AI landscape.
To this end, the study considers the top ten universities in 2024 according to QS World University Rankings, Academic Ranking of World Universities and Times Higher Education World University Rankings. Since most of them appeared in all three rankings, research has considered each institution once, delimiting a corpus of 16 institutions. The analysis employs an inductive content analysis methodology, which permits the examination of publicly available material on university websites. Texts were fully read, segmented into meaningful units, and inductively labelled and categorized. The coding process started from an individual analysis and evolved into a collaborative phase, leading to a shared coding system.
Preliminary findings indicate that (Gen)AI guidelines undertake multiple actions addressed to all academic community members. Faculty and staff play a key role. Teachers act as primary disseminators, designing course-specific guidelines incorporated into curricula and sharing them with students. Furthermore, guidelines sustain all stages of (Gen)AI adoption. Before its usage, universities establish preventive and informational actions to foster awareness and reflection, enhancing teaching and learning activities. Institutions provide specific procedures and warnings for its implementation, clarifying mandatory and forbidden actions. Finally, universities stress the principle of responsibility and ownership as after-usage actions. However, many existing guidelines remain general, allowing broad interpretation. While they offer a valuable tool for understanding (Gen)AI, they do not always fit every context or offer the most optimal solutions. The research stresses the actual and potential value of guidelines but acknowledges they are only one of the several ways to cope with (Gen)AI.
By providing a clearer picture of current developments, the research contributes to theoretical and practical advancements, offering a foundation for future inquiries into the role of (Gen)AI in Higher Education. Future work should focus on reducing ambiguity and enhancing the applicability of these guidelines. By doing so, institutions can foster a more structured and informed approach to integrating (Gen)AI into teaching and learning.
Keywords: (Gen)AI Guidelines, Higher Education, Teaching and Learning, Ethics.