Students’ Perceptions of Ai-Assisted Feedback in Speaking Practice of Higher Education Students

Abstract

The integration of Artificial Intelligence (AI) in language education has transformed how students receive and engage with feedback, particularly in developing speaking proficiency. This study explores university students’ perceptions of AI-assisted feedback in English-speaking practice within an Indonesian higher education context. Using a qualitative descriptive design, data were collected from semi-structured interviews and reflective journals of 15 EFL students who regularly used AI-based tools such as ELSA Speak, SpeechAce, and ChatGPT Voice during speaking exercises. Thematic analysis revealed three major findings: (1) students perceived AI feedback as immediate, consistent, and helpful for improving pronunciation and fluency; (2) AI-based interaction enhanced learner autonomy and speaking confidence by providing a safe, nonjudgmental practice space; and (3) participants expressed skepticism toward the accuracy and communicative depth of AI-generated feedback, noting that it often neglected discourse and pragmatic aspects of speaking. Overall, students viewed AI-assisted feedback as a complementary pedagogical aid rather than a substitute for teacher evaluation. The findings underscore the importance of human–AI collaboration in language teaching, suggesting that AI tools are most effective when integrated with teacher mediation and reflective guidance. This study contributes to the growing body of literature on technology-enhanced speaking instruction and provides insights for educators seeking to implement AI tools responsibly and effectively in higher education contexts.