Аңдатпа
This study aimed to establish conceptual mechanisms and patterns for applying artificial intelligence (AI) technologies to personalised physics education within the Kazakhstani educational system. Methods. A comprehensive methodology was employed, combining both theoretical and empirical approaches. The theoretical phase involved reviewing regulatory documents in Kazakhstan’s education system. The empirical phase consisted of a pedagogical experiment conducted between September and December 2024 across three educational institutions in Taldykorgan, Kazakhstan. The study involved 58 tenth-grade students, divided into experimental and control groups; the experimental group utilised an AI-driven personalised learning system designed to adapt content based on student performance. Data were collected on academic performance, theoretical knowledge, practical skills, and research competencies using pre-tests, interim assessments, and final evaluations. Results. The experimental groups demonstrated a significant improvement in academic performance, with the average score increasing from 4.2 to 4.6. 76% of students in experimental groups successfully solved advanced problems, compared to 52% in control groups. The system fostered improved critical thinking, research competencies, and self-assessment skills, while enhancing students’ ability to engage in scientific discourse and apply knowledge in interdisciplinary contexts. Conclusions. The AI-driven model proved highly effective at personalising learning and improving students' academic and cognitive outcomes. It offers a scalable framework for adapting content to individual learning styles, with positive impacts on motivation and problem-solving abilities. The findings contribute to a growing body of research on AI applications in education and provide a foundation for further advancements in personal learning systems.