Critical analysis of the technological affordances, challenges and future directions of Generative AI in education: a systematic review

　　This systematic review synthesizes 27 core-journal papers (2020–2023) on generative AI in education using a PRISMA-guided selection and inductive thematic coding. It maps technological affordances, key challenges, and actionable directions for future research and practice.   Highlights Four technological affordances: Accessibility (always-on support; remote learning), Personalization (context-aware feedback/materials), Automation (offloading repetitive tasks; boosting preparation/assessment), and Interactivity (AI as conversational partner supporting language and conceptual learning). Five central challenges: Academic integrity (plagiarism/cheating), Response errors & bias (hallucinations, fabricated citations, data bias), Over-dependence (risks to higher-order thinking), Digital divide (paywalls and bans), and Privacy & security. Roles in educational settings: Generative AI can function as an intelligent tutor, tutee, learning tool/partner, and domain expert—supporting curriculum design, learning assistance, and teacher PD. Fig 1. Roles of generative AI in education   Methods 　　The review searched Web of Science, Scopus, and ScienceDirect, and coded contexts, sectors, and methods to portray the research landscape and use-cases. Fig 2. Guidelines about the future directions of research and practice   Conclusions and Recommendations Assessment & policy: Adopt diverse assessments (proctored open-ended tasks, orals, process portfolios) and enhance misconduct detection; establish and maintain institutional AI ethics guidelines. Bias mitigation & capacity building: Continuously monitor model bias and equip educators with AI literacy (including prompt design) to ensure human–AI collaboration that improves material design and feedback. Curricular integration: Integrate Generative AI into learning activities to foster higher-order skills; position AI as a thinking aid rather than an answer engine, with robust data protection and informed consent.   Contribution 　　The review bridges role-based perspectives with technological affordances to present a balanced account of opportunities and risks, offering concrete guidance for policymakers, educators, and learners.