Despite the growing interest in artificial intelligence for language education, empirical evidence regarding the use of AI tools like ELSA Speak in senior high school classrooms remains limited, especially in rural Indonesian settings. To address this issue, this study investigates how integrating the ELSA Speak app into regular English instruction affects EFL students' speaking abilities. The research utilized a quasi-experimental pre-test and post-test control group design involving 50 twelfth-grade students at SMAN 01 Luwu. Half of the students learned using the app (experimental group), while the rest followed standard teaching methods (control group). Based on the Mann–Whitney U test, the students who used ELSA Speak achieved significantly higher post-test scores than those in the control group (U = 483.00, p = 0.001, r = 0.46). Specifically, the AI-supported learning process led to major improvements in the students' pronunciation, fluency, and speaking confidence. In conclusion, this study strengthens the existing literature on Mobile-Assisted Language Learning (MALL) by showing that combining AI applications with traditional teaching can successfully solve the lack of speaking practice and enhance specific oral language skills in high school environments.
Copyrights © 2026