<![CDATA[While artificial intelligence (AI) technologies such as conversational chatbots and speech recognition systems are increasingly adopted in language education, existing research focuses predominantly on isolated tools rather than integrated instructional models. This study proposes and evaluates a Human–AI Collaborative Blended Learning Framework for the development of English as a Foreign Language (EFL) speaking skills. Using a two-phase research design, the study first synthesizes contemporary literature to establish a 7-layer conceptual framework that connects stakeholder participation, learner profiling, a Human–AI collaboration engine, blended activities, adaptive speaking services, teacher facilitation, and continuous learning analytics. Second, the framework was evaluated through an expert validation survey involving 52 specialists in language education, applied linguistics, and educational technology (40.4% PhDs, 57.7% Master’s). Quantitative findings revealed strong endorsement for actionable teacher support dashboards (80.8% agreement), visual pronunciation feedback (78.8%), and private AI speaking practice for anxiety reduction (75.0%; M = 2.98/4.00). However, experts expressed reservations regarding automated prompts for language elaboration (53.8% agreement), fixed 14-meeting schedules (57.7%), and automated grading weights (57.7%). Qualitative analysis highlighted that acceptance remains conditional upon accent fairness, data privacy, and active teacher oversight. Implications: Theoretically, the study advances AI-assisted language learning by establishing a human-centered collaborative model that prioritizes teacher-AI shared regulation over technological substitution. Pedagogically and practically, the findings imply that institutions and EdTech developers should deploy AI primarily for scalable, low-stakes rehearsal and actionable lesson planning, while preserving teacher authority over social communication, pragmatic competence, and high-stakes assessment.]]>
Copyrights © 2026
