Impact of Generative Artificial Intelligence on Education

Abstract

Generative artificial intelligence (GenAI) has quickly become part of education through tools that can create text, explanations, code, images, feedback, and other learning resources. This paper studies how GenAI affects education by reviewing recent work on teaching, learning, assessment, academic integrity, accessibility, teacher practice, and institutional governance. A structured literature review was used to identify and group peer-reviewed studies and major policy guidance published during the emergence of contemporary GenAI, with more attention to higher education while also considering schools. The reviewed studies show that GenAI can support personalized learning, rapid formative feedback, brainstorming, language development, content generation, tutoring, and teacher productivity. Recent meta-analyses also report benefits for several university learning outcomes, but the results differ depending on learner, tool, task, intervention design, and context. However, hallucinated information, bias, privacy risks, unequal access, overreliance, reduced metacognitive effort, authorship ambiguity, and assessment disruption create important challenges. The paper presents a human-centred framework in which pedagogical purpose, AI literacy, verification, assessment redesign, equity, privacy, and institutional governance together shape the educational value. The paper concludes that GenAI should be treated neither as a replacement for teachers nor as a universally beneficial shortcut, but as an assistive technology whose value depends on responsible pedagogical integration and continued research and evaluation.