FROM FRAMEWORK TO PRACTICE: A PEDAGOGICAL MODEL FOR DESIGNING AI-INTEGRATED ENGLISH LESSONS FOR SCHOOL AND UNIVERSITY LEARNERS

Abstract

Abstract: Building directly on the authors’ methodological framework, which mapped five categories of artificial intelligence (AI) tools onto specific EFL speaking and writing sub-skills and consolidated a five-stage AI-integrated instructional cycle [10], the present thesis addresses the practical question that framework left open: how a classroom teacher should organize a lesson, allocate time, and select tools so that AI use becomes a quality-controlled pedagogical event rather than an incidental add-on. Using a qualitative, literature-based design synthesis of implementation-focused sources published between 2021 and 2026, the study translates the five-stage cycle into a concrete, time-structured lesson blueprint, differentiated for school-age (grades 5-11) and university EFL learners, and proposes a set of lesson-quality criteria that teachers and methodologists can use to evaluate AI-integrated lessons. The analysis shows that classroom-implementation feasibility, and not pedagogical effectiveness alone, differs systematically between school and university settings: speech-recognition tools are comparatively easy to deploy with younger learners, whereas generative AI and natural-language-processing (NLP) writing checkers require a level of learner autonomy and device access more reliably available at university level. The proposed model organizes each lesson stage around explicit teacher actions, learner actions, AI-tool choices, and formative-assessment checkpoints, and is illustrated with a sample eighty-minute university-level lesson plan and a level-differentiated adaptation table. The discussion addresses teacher AI-literacy preparation, equitable access, and safeguards against learner over-reliance on automated correction, and offers concrete recommendations for methodologists designing AI-integrated EFL curricula within school and university systems in Uzbekistan.