Introduction. The article examines the factors influencing students’ reflective skills in AI-based learning in higher education. Artificial intelligence (AI) has become an important component of the digital transformation of
education, providing opportunities to enhance efficiency, accessibility, and flexibility of learning. In this context,
the development of students’ reflective and metacognitive skills has become an increasingly important research area. The study aimed to identify key factors associated with students’ reflective skills in AI-based learning environments and to empirically examine the relationships among these factors. Methodology and Methods. The study employed quantitative research design and Partial Least Squares Structural Equation Modeling (PLSSEM). The sample comprised 234 students from Abai Kazakh National Pedagogical University and Kazakh National Women’s Teacher Training University. The model included four factors: perceived usefulness of AI, perceived ease of use, AI usage, and AI feedback. The results revealed a weak positive relationship between
perceived usefulness of AI and students’ reflective skills, while perceived ease of use, AI usage, and AI feedback
demonstrated weak negative relationships. These findings indicate that AI use alone does not ensure the
development of students’ reflective thinking. Effective integration of AI into higher education should therefore
be supported by reflective tasks, critical thinking activities, and metacognitive strategies. Scientific novelty.
The study empirically examines the relationships between AI-related learning factors and students’ reflective skills, demonstrating the limited explanatory value of technology use alone and highlighting the importance of pedagogical and metacognitive conditions. Practical significance. The findings support the integration of reflective tasks, critical thinking, self-assessment, and metacognitive activities into AI-supported learning environments. The study provides a basis for developing pedagogically grounded approaches to the effective integration of AI technologies in higher education.