<![CDATA[This study examines how the use of ChatGPT, an AI-driven chatbot, influences learners’ speaking anxiety and their spoken performance in an English as a Foreign Language setting. The analysis is guided by a framework of language-learning anxiety that highlights three key components of speaking anxiety: communication apprehension, fear of negative evaluation, and test anxiety. A qualitative approach was employed, involving 30 university students from Eastern Indonesia. Data were collected through a 25-item Likert-scale questionnaire adapted from FLCAS, six open-ended questions, and semi-structured interviews with 10 selected participants. The findings reveal that ChatGPT plays a crucial role in reducing communication apprehension, fear of negative evaluation, and test anxiety, allowing students to practice speaking English in a low-pressure environment. Notably, findings showed that after using ChatGPT in speaking practice, 60% of students experienced moderate speaking anxiety, while 26.67% exhibited high levels and 13.33% reported low anxiety. Additionally, frequent AI-assisted interaction enhances fluency, confidence, and structured speech formulation, improving oral proficiency. However, limitations were identified in pronunciation feedback, conversational naturalness, and expressive interaction, suggesting that AI should serve as a supplementary tool rather than a standalone language learning solution. Given these findings, future AI improvements should focus on voice recognition for pronunciation accuracy, adaptive conversational dynamics, and multimodal interaction strategies. This study underscores ChatGPT’s potential to effectively reduce speaking anxiety while advocating for integrated AI-human interaction approaches in language education to support comprehensive and immersive speaking practice.]]>
Copyrights © 2025
