Artificial Intelligence in English Language Teaching: Applications, Benefits, and Challenges — A Critical Integrative Review

Abstract

Artificial intelligence (AI) offers opportunities for feedback, spoken interaction, reading support, and listening practice in English language teaching (ELT). This critical integrative review examines selected scholarship to distinguish AI capabilities, learner perceptions, task performance, and evidence of independent language development. Ten purposively chosen illustrative publications were categorized by their primary language-skill focus: six writing (60%), two speaking (20%), one reading (10%), and one listening (10%). Writing or speaking therefore accounted for eight of these ten examples (80%). These proportions describe only this selected illustrative set, not the prevalence of skills in ELT research as a whole. The literature highlights possibilities for feedback, rehearsal, and task-specific support alongside concerns about accuracy, overreliance, academic integrity, privacy, and unequal access. An evidence-to-claim framework is proposed to connect AI function, instructional task, measured outcome, and limitations of inference. The review does not claim a systematic database search or a PRISMA-screened study sample. Future systematic studies should use archived searches, explicit eligibility decisions, and quality appraisal to test the broader distribution and effectiveness of AI applications in ELT.