Answers are grounded in this paper and cite their sources.
The integration of artificial intelligence into physics education raises questions that are fundamentally epistemological, not merely technological. As generative AI systems increasingly produce explanations, representations, and solution pathways, they begin to participate in the very practices, modelling, inquiry, and reasoning, that constitute the disciplinary core of physics learning. This perspective article presents a theoretically grounded argument, drawing on established physics education research (PER), that the educational value of AI in STEM classrooms depends not on its technical capabilities but on how it is epistemically framed within instructional activity. We examine three interconnected dimensions: AI’s role as a collaborator in physics modelling and inquiry, the reconfiguration of teacher authority and professional knowledge, and the implications for student agency and epistemic justice. We argue that when AI is positioned as a provisional contributor whose outputs are subject to physical reasoning, testing, and revision, it can strengthen disciplinary sense-making. However, when treated as an authority or shortcut, it risks displacing the meaning-making processes on which conceptual understanding depends. The article concludes by outlining implications for physics education research, instructional design, and teacher preparation in AI-integrated STEM environments.
No takes yet. Share an insight, caveat, or question.
Kotsis Konstantinos T. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: