This study was conducted to systematically review Korean research on the use of generative artificial intelligence in physical education and to synthesize research trends, educational applications, and implementation issues. Korean journal articles published from January 2020 to December 2025 were collected from RISS, KISS, DBpia, and KCI. A PICO-informed framework guided the search and selection process, and 17 studies were included after duplicate removal, screening, and full-text review. Data were extracted using a standardized form, and findings were inductively categorized with reference to Strauss’s (1987) open-coding procedures. The results were organized into four categories: instructional application approaches, learning outcomes and educational effects, teacher factors, and implementation conditions and issues. Overall, generative AI was used mainly to support instructional design, cognitive understanding, and reflective learning rather than to replace physical education lessons. However, studies treating generative AI as the central intervention and addressing implementation conditions such as data-based assessment and feedback, ethics, and infrastructure remained limited. Future research was suggested to refine physical education–specific instructional design principles and to strengthen integrated systems for assessment, feedback, and teacher professional development.
Seung-Bo Sim (Sat,) studied this question.