Answers are grounded in this paper and cite their sources.
Since the release of ChatGPT, AI assistants based on Large Language Models (LLM) have quickly made their way into higher education. Whether this new generation of chatbots can be used to effectively drive learning processes, however, depends on an equally new digital skill: prompt engineering. Considering both educational commitments and legal restrictions, this article addresses the questions of why and how students in higher education, especially in the European Union, need to be introduced to prompt engineering if they are to successfully integrate LLM-based chatbots into their self-study processes. Against the backdrop of the Technology Acceptance Model (TAM) and existing research on the factors influencing university students’ AI use, this article presents a quadrant model meant to help educators and researchers assess the characteristic skills and needs of four different LLM user groups within the student body: the AI-confident Experts, the LLM-shy Avoiders, plus the Improvisers and Independents who fall somewhere in between these extremes. Focusing on students’ AI-related learning anxiety and learning self-efficacy as two factors that influence LLM adoption, we discuss how these four target groups might be best supported, both individually and as a heterogeneous student body.
No takes yet. Share an insight, caveat, or question.
Schelling et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: