Ai-driven Approaches To Advance Speaking Proficiency In Lmoocs: Insights, Innovations, And Pedagogical Implications

LMOOCs, or Language Massive Online Open Courses, which are a fairly novel contribution to the plethora of online resources readily accessible for language instruction, have been permeating academic institutions in recent years. Regardless, concerns over developing the speaking skill of a substantial demographic of EFL learners on LMOOCs persist hitherto. Relative to other linguistic skills, speaking necessitates real-time interaction, personalised feedback, and sustained engagement to ensure enhancement; the very elements that are challenging to attain given the abundant numbers of LMOOC participants compared with limited instructor availability. The significance of this research lies in its attempt to orient instructors towards the most effective Artificial Intelligence (AI) tools to enhance speaking proficiency in online learning settings; by extension, transfer them into LMOOC context. The current research aims in unveiling AI tools tailored to key speaking sub-skills (linguistic competence, oral fluency, interaction, and production) found in the growing volumes of research literature while also addressing core issues such as learner motivation, automated feedback, and interactive speaking practice. Adopting a systematic literature review approach, this research examines a total of (N=53) papers from four major academic databases (Scopus, ScienceDirect, JSTOR, and SpringerLink). The span of the extracted data for the systematic literature review covers 5 years, from 2019 to 2024. The results provide teachers and course designers with a road map for creating more dynamic, responsive, and effective LMOOC experiences for EFL learners to solve problems related to learner engagement, personalised feedback, and interactive practice through incorporating AI tools. They also offer actionable insights into AI integration strategies into LMOOC framework.