As artificial intelligence (AI) becomes increasingly present in higher education, students’ perceptions of AI shape how they engage with these technologies. This paper presents a small-scale exploratory study examining how university students visually conceptualise AI through drawings. It aims to identify emerging patterns and their relation to broader digital representations of AI.

The dataset comprises 30 drawings produced by small groups of English language students at a Finnish university, collected with informed consent. Using qualitative visual content analysis, the study combines inductive coding with attention to recurring motifs, symbols, and compositional features (Rose, 2016). The analysis explores how abstract concepts such as intelligence and agency are visually constructed and draws on arts-based methods that critique GenAI technologies (Lupton, 2026).

The findings reveal a dominance of familiar visual patterns: many drawings depict AI as humanoid robots, disembodied brains, or network-like structures, echoing common stock images online. This aligns with research demonstrating that widely used stock images of AI, such as humanoid robots and glowing brains, can reinforce simplified or inaccurate understandings of the technology (Dihal & Duarte, 2023). A smaller subset of drawings diverges from these patterns and portrays AI as embedded in everyday contexts or as a collaborative partner in learning and work. These alternative representations align with perspectives that frame AI as a relational and socio-technical phenomenon rather than purely autonomous (Long & Magerko, 2020).

By visualising students’ underlying assumptions and imaginaries, this study contributes to discussions on human–AI collaboration in education. The findings suggest that dominant visual narratives may limit understanding and reinforce narrow conceptions. As a work in progress, the paper invites reflection on methods for analyzing visual data and the pedagogical implications of critically engaging with AI imagery in higher education.