<![CDATA[Global Englishes Language Teaching (GELT) promotes inclusivity by valuing diverse English varieties, but GenAI systems often default to native-speaker norms, risking linguistic homogenization. This study investigates how GenAI can be integrated into GELT to authentically represent diverse Englishes, and what ethical and psychological challenges accompany its adoption in classrooms. A qualitative multi-case study was conducted across three Indonesian universities, combining classroom observations, teacher and student interviews, institutional policy analysis, surveys, and digital interaction logs. Data were examined through thematic coding, critical discourse analysis, narrative inquiry, and descriptive statistics. Each case was interpreted by identifying context-specific themes and discursive patterns, while the cross-case analysis compared these findings to highlight convergences, divergences, and broader ethical implications across language education settings. Findings show that while GenAI supports material development and sparks critical classroom dialogue, its representations are frequently tokenistic and reinforce deficit views of non-native varieties. Teachers mediate these biases by critiquing outputs and affirming student linguistic identities, though institutional policies emphasizing standard English limit such efforts. Ethical and psychological challenges include student over-reliance, reduced autonomy, teacher workload, and inequities in access. The study concludes that AI integration is an ecological and pedagogical issue, requiring critical teacher agency, student empowerment, and institutional policies that align innovation with linguistic justice.]]>
Copyrights © 2026
