iSchool students’ mental models of GenAI
Irene Lopatovska, Conor Mack, Ellen Connors · Information Research · 2026
Introduction. Knowledge of users’ mental models (an understanding of a system and its functions) helps align system behaviour with user expectations and identify user concerns that influence adoption. We examined graduate students’ mental models of generative artificial intelligence (GenAI) as part of a larger study of GenAI adoption for academic purposes at the Pratt Institute iSchool. Method. The data about mental models of GenAI were collected through drawings and conversations with participants during focus groups. Analysis. The qualitative visual and textual data were analysed using thematic analysis. Results. Students’ mental models reflect a high-level understanding of the technical complexity of the GenAI system, with occasional acknowledgments of end-users and collection creators, as well as ethical, environmental, and quality concerns. mental models often reflect knowledge of prior information technology and use human analogies to explain the GenAI. Conclusions. Even a technologically literate group of information science graduate students exhibits limitations in understanding GenAI systems. iSchool students are not only avid users of GenAI—whose deeper understanding of this technology could enhance usage efficiency and satisfaction—but also future designers of GenAI systems. Technology features, additional experiences with technology, educational resources/curricula, and critical media coverage could improve their understanding of GenAI’s technological and social dimensions.