Harnessing AI for Pronunciation Development: Investigating the Impact of Elsa Speak on EFL Learners’ Oral Proficiency

Abstract

Pronunciation plays a pivotal role in oral fluency and overall communicative effectiveness in English as a Foreign Language (EFL) learning. However, pronunciation instruction remains a persistent challenge in higher education contexts, particularly where learners have limited opportunities for individualized feedback and sustained oral practice. Responding to recent advances in artificial intelligence (AI), this study investigated the effectiveness of the AI-powered ELSA Speak application in enhancing EFL learners’ pronunciation development. Adopting a longitudinal experimental design, the study involved 60 university-level EFL students, equally divided into experimental and control groups. Learners’ pronunciation performance was assessed at three time points using the CAFIS analytic framework. Repeated-measures analyses revealed that students using ELSA Speak demonstrated significantly greater improvement across all CAFIS dimensions compared to the control group. Accuracy showed the most immediate gains, followed by gradual but consistent development in intonation and fluency over time. These findings underscore the value of AI-assisted pronunciation instruction and highlight the importance of adopting multidimensional assessment frameworks to capture nuanced developmental changes. The study contributes empirical evidence from a Saudi higher education context and offers pedagogical insights into the integration of AI-driven tools for pronunciation development in EFL classrooms.