Generative AI in Physics Education: A PRISMA-Guided Systematic Review of Empirical Studies

Abstract

Generative artificial intelligence (GenAI) tools are redefining education, particularly in science disciplines such as physics. Even though these systems are widely used, there are still only a few studies that clearly summarize their use in teaching and their impact. In this PRISMA-guided systematic review, we analyzed 21 empirical studies (from 2023 to 2025) on the role of GenAI in physics education. We combined bibliometric mapping with content-based refinement and classified the studies by educational level, tools used, research methodology, and pedagogical role. Three main functions were identified: problem solver (57% of studies), student’s tutor (38%), and teacher’s assistant (33%). Problem-solving roles demonstrated variable accuracy, ranging from 100% to 0% across task types. Tutoring roles, though few, showed the strongest evidence of positive learning outcomes. Assistant roles mainly supported efficiency in grading and content generation, with less direct evidence of learning outcomes. The review contributes by providing this physics-specific PRISMA synthesis of GenAI in education and by identifying the roles GenAI tools assume in this context. This research is limited by the small, diverse set of included studies, which focus mainly on university contexts, and by its reliance on Scopus-indexed literature. Our findings suggest that GenAI tools can potentially improve conceptual understanding, formative assessment, and instructional efficiency. However, their effective integration requires human supervision, critical evaluation, and strategies to address issues of accuracy, equity, and multimodality.

Received: 12 August 2025 | Revised: 29 December 2025 | Accepted: 28 January 2026

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Author Contribution Statement

Maria Moundridou: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Supervision. Nikolaos Moutis: Conceptualization, Methodology, Writing – review & editing, Supervision, Project administration. Konstantinos Charalampopoulos: Investigation, Writing – review & editing. Nikolaos Matzakos: Conceptualization, Methodology, Formal analysis, Data curation, Writing – original draft, Writing – review & editing, Visualization.