Teachers’ perspectives on pedagogical challenges of AI integration in English language teaching

Abstract

Artificial intelligence (AI) tools are rapidly being integrated into English language teaching (ELT). However, limited empirical evidence exists regarding the pedagogical challenges teachers encounter during classroom implementation. This study investigates whether demographic variables influence ELT teachers’ perceptions of AI effectiveness and related pedagogical challenges. A quantitative survey design was employed, involving 200 ELT teachers with prior AI exposure. Data were collected using a validated Likert-scale instrument (α=0.81) and analyzed using independent samples t-tests, one-way ANOVA with Tukey post hoc analysis, and Chi-square (χ²) tests. Results revealed significant gender differences in perceived effectiveness (t=2.504, p=0.044) and pedagogical challenges (t=2.627, p=0.018), with female teachers reporting higher mean scores. Locality differences were significant only for perceived effectiveness (F=4.610, p<0.05), particularly between urban and rural teachers. No significant associations were found for age (χ²=12.830, p>0.05) or teaching experience (χ²=4.266, p>0.05). Educational level taught was significantly associated with perceived effectiveness (χ²=13.816, p<0.05). The findings suggest that AI integration challenges are largely systemic, even though perceptions of effectiveness vary across contexts.