Exploring Mathematics Teachers’ Perceptions of ChatGPT for Personalized Learning and Technology Dependence

The development of artificial intelligence, particularly ChatGPT, has begun to influence mathematics learning in classrooms. This technology is not only used as a learning support tool but also affects how teachers teach and how students understand mathematical concepts.  This study aims to analyze the effectiveness of ChatGPT in mathematics learning, its benefits and challenges, and the potential for technology dependence based on mathematics teachers’ perspectives. This study employed a Systematic Literature Review (SLR) approach. The data analysis technique used the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) method through the stages of identification, screening, eligibility, and article inclusion. Articles were collected from ScienceDirect, Google Scholar, Semantic Scholar, Publish or Perish, and ResearchGate. Based on the selection process, 27 articles were analyzed thematically according to the research focus. The findings indicate that mathematics teachers perceive ChatGPT as effective in supporting more flexible, interactive, and adaptive learning based on students’ needs. ChatGPT also supports personalized learning through adjusted explanations and learning pace. However, its use also presents challenges related to information accuracy, teachers’ readiness, and supervision of AI use. In addition, the ease of obtaining instant answers may increase technology dependence and reduce students’ engagement in mathematical thinking during the learning process.