DEVELOPING EFL LEARNERS' ORAL SPEECH COMPETENCE THROUGH METACOGNITIVE STRATEGY INSTRUCTION SUPPORTED BY ARTIFICIAL INTELLIGENCE

The rapid development of artificial intelligence (AI) has created new opportunities for foreign language education, particularly in the development of learners’ oral communication. However, access to AI-mediated interaction alone does not automatically guarantee the development of sustainable speaking competence. Learners also need to understand how they plan, monitor, evaluate, and regulate their own speaking performance. This article examines a methodological framework for developing English as a Foreign Language (EFL) learners’ oral speech competence through the systematic teaching of metacognitive strategies supported by artificial intelligence. The theoretical foundation combines metacognition theory, language learning strategy research, communicative approaches to speaking, and recent research on AI-assisted language learning. The methodology of the article is based on a critical synthesis of established theoretical works, systematic reviews, meta-analyses, and empirical studies. Particular attention is given to planning, monitoring, evaluation, self-correction, goal setting, reflection, and strategy transfer. Recent research indicates a rapid expansion of AI-supported EFL research. A 2026 systematic review examined 708 publications from 2019–2024 and retained 103 studies, while a 2024 systematic review specifically examining AI chatbots for EFL speaking identified 24 empirical studies published between 2017 and 2023. A 2025 meta-analysis synthesized 23 experimental and quasi-experimental studies involving 1,872 participants and reported a statistically significant overall effect of AI-based interventions on EFL outcomes (Hedges’ g = 1.10, 95% CI [0.76, 1.46], p < .001), although heterogeneity was high (I² = 92.66%). These findings support the potential value of AI while simultaneously demonstrating the importance of pedagogical design. The article proposes a three-stage AI-supported metacognitive cycle—planning, monitored performance, and evaluation/regulation—as a practical methodology for EFL speaking instruction. The central argument is that AI should function as a scaffold for metacognitive development rather than as a substitute for teachers, human interaction, or learner cognitive activity.