Artificial intelligence (AI) tools have the potential to revolutionise educational practices, but their use raises ethical and organisational concerns for higher education institutions (HEIs). We investigated Italian students' perception and usage of AI tools at the University of Udine using questionnaires (N = 531) with fixed and open-ended items and semi-structured interviews (N = 60). The findings reveal that while over two-thirds of students use ChatGPT for personal purposes, 40.9% employ it for academic tasks, primarily for summarisation, clarification and content generation. Most students (70%) expressed awareness of ethical concerns, and 55% saw the potential value in using AI, if regulated, for specific learning activities. Students generally did not perceive a direct threat to their studies or career prospects, although 60% voiced reservations about potential negative impacts on broader societal functions. These results underscore the need for HEIs to implement regulations for AI tool use to balance innovation with ethical challenges.