User profiles and perceptions of artificial intelligence in higher education physics

Abstract This study investigates how undergraduate physics students at a public university in Barcelona (Spain) use GenAI tools in their academic work, examining patterns of use, perceptions, and implications for teaching methods. Through a collaboration between physics faculty and science education researchers, a mixed-methods approach was employed, including surveys ( n = 172), cluster analysis, and focus groups. As a result, five distinct user profiles were identified, ranging from intensive and varied artificial intelligence (AI) users to those maintaining a critical distance from AI. These five clusters reflect diverse motivations, from accelerating routine tasks to enhancing conceptual understanding, alongside varying degrees of trust and engagement with AI. These findings highlight the need to reconsider learning objectives and assessment strategies within physics curricula, promoting the responsible integration of AI as an educational ally while addressing ethical and cognitive challenges.