Culturally adaptive AI in BIPA classrooms: Learner voices and pedagogical implications

¹ Department of Indonesian Language and Literature Education, Universitas Negeri Surabaya, Indonesia
² School of International Letters and Cultures, Arizona State University, USA
³ Department of Language and Arts, Maejo University, Thailand
⁴ Department of Graphic Design, Universitas Negeri Surabaya, Indonesia
⁵ Department of Education Technology, Universitas Negeri Surabaya, Indonesia
⁶ Department of English Language Education, Universitas Islam Kadiri, Indonesia
⁷ Department of Indonesian Language and Literature Education, Universitas Negeri Surabaya, Indonesia

^* Corresponding author: primaasteria@unesa.ac.id

Abstract

This study explores the pedagogical, ethical, and cultural implications of using generative artificial intelligence (AI) in BIPA (Bahasa Indonesia bagi Penutur Asing) classrooms. Drawing insights from five international learners two from Korea, two from China, and one from Malaysia the research investigates how AI tools such as ChatGPT and Gamma influence speaking performance, learner confidence, and cultural adaptation. While AI-supported learning facilitated grammatical fluency and reduced anxiety, participants reported challenges including overformal language, lack of colloquial expressions, and cultural misalignment. Methodologically, the study employed a qualitative interpretive approach using interviews, classroom observations, and student-produced materials, we analyzed both the benefits and the limitations of these tools. Findings shows the importance of designing AI that is transparent, culturally sensitive, and guided by ethical teaching practices. We hope this research helps educators and developers build AI systems that support not replace human-centered, inclusive, and adaptive language learning. The study contributes to the discourse on inclusive language pedagogy and aligns with the objectives of SDG 4 (Quality Education) and SDG 10 (Reduced Inequalities).