How ChatGPT Helps the Education Sector — A Comprehensive Review

AbstractLarge language models (LLMs) such as ChatGPT have rapidly entered educational practice. This review synthesizes empirical findings and systematic reviews on ChatGPT’s use in education, summarizing core benefits (personalized tutoring, content generation, writing support, scalable feedback), persistent risks (hallucinations, academic-integrity threats, equity and privacy issues), and recommended institutional and pedagogical responses. Evidence from systematic reviews and large meta-analyses indicates that, when used carefully, ChatGPT can improve learning outcomes and formative experiences; however, major unresolved problems (detection, overreliance, cognitive effects) require policy and instructional changes. Keywords: Large language models; ChatGPT; artificial intelligence in education; personalized learning; academic integrity; formative feedback; educational policy