The Effect of AI Chatbots on Students’ Oral Communication Proficiency

This study examined the effectiveness of integrating Artificial Intelligence (AI) chatbots in enhancing the oral communication proficiency of Grade 11 Humanities and Social Sciences (HUMSS) students. Anchored in an explanatory sequential mixed-methods design, the research combined quantitative pretest–posttest measures with qualitative data from focus group discussions to provide a comprehensive analysis of learning outcomes and student experiences. Over a nine-week intervention, students participated in AI-assisted oral communication tasks embedded in GRASPS-based lessons aligned with the K–12 Oral Communication curriculum. Conversational AI tools, including ChatGPT, Bing Chat, and Copilot, were utilized to simulate real-life speaking situations and provide real-time, low-pressure interaction opportunities. Quantitative findings revealed a statistically significant improvement in students’ oral proficiency scores, particularly in fluency, coherence, vocabulary use, and communicative confidence. However, increased posttest variability highlighted the need for differentiated support. Qualitative findings further indicated that students perceived AI chatbots as nonjudgmental and supportive conversational partners, which reduced anxiety and encouraged risk-taking in speaking, supporting Krashen’s Affective Filter Hypothesis and Vygotsky’s concept of scaffolded learning within the Zone of Proximal Development. Despite these benefits, challenges such as unstable internet connectivity, limited device access, and the need for guided pedagogical integration were identified. Overall, the study concludes that AI chatbots, when strategically and ethically integrated alongside teacher facilitation and reflective activities, can effectively complement traditional instruction and foster a more inclusive, student-centered environment for developing oral communication skills.