The impact of Generative AI (GenAI) on practices, policies and research direction in education: a case of ChatGPT and Midjourney

TLDR

Generative AI tools are increasingly accessible and are reshaping school education, yet most discourse has focused on higher education, underscoring the need to understand teachers’ perspectives. This qualitative study investigates how GenAI alters school education from the viewpoints of teachers and school leaders. The authors surveyed 88 teachers and leaders and conducted focus groups using a four‑domain framework—learning, teaching, assessment, and administration—to examine the effects of ChatGPT and Midjourney. Thematic analysis revealed four main themes and 12 subthemes, yielding practice recommendations (e.g., addressing know‑it‑all attitudes, prerequisite knowledge, interdisciplinary teaching), policy implications (new assessment, AI education, professional standards), and six future research directions.

Abstract

Generative artificial intelligence (GenAI) tools have become increasingly accessible and have impacted school education in numerous ways. However, most of the discussions occur in higher education. In schools, teachers’ perspectives are crucial for making sense of innovative technologies. Accordingly, this qualitative study aims to investigate how GenAI changes our school education from the perspectives of teachers and leaders. It used four domains – learning, teaching, assessment, and administration – as the initial framework suggested in a systematic literature review study on AI in education. The participants were 88 school teachers and leaders of different backgrounds. They completed a survey and joined a focus group to share how ChatGPT and Midjounery had a GenAI effect on school education. Thematic analysis identified four main themes and 12 subthemes. The findings provide three suggestions for practices: know-it-all attitude, new prerequisite knowledge, interdisciplinary teaching, and three implications for policy: new assessment, AI education, and professional standards. They also further suggest six future research directions for GenAI in education.