Technostress forecasting in EFL classrooms: How digital fatigue and perceived AI ethical concerns shape digital burnout and undermine SDG-4 achievement

This study examines how digital stressors technostress, digital fatigue, and perceived AI ethical concerns impact foreign language learning anxiety (FLLA), digital burnout, and academic performance in English as a Foreign Language (EFL) classrooms, with implications for achieving Sustainable Development Goal (SDG) 4 on quality education. Employing a cross-sectional design, data were collected from 545 university-level EFL learners (domestic and international) in China. Partial least squares structural equation modeling (PLS-SEM) was used to test a serial mediation model with moderation. Results show that technostress (β = 0.41), digital fatigue (β = 0.33), and AI ethical concerns (β = 0.22) significantly increase FLLA, which in turn fully mediates the effects of these stressors on digital burnout (indirect effects: 0.10-0.19). Technology self-efficacy moderates the negative relationship between burnout and academic performance (β = 0.21), with this moderating effect being stronger for international students. Academic performance strongly predicts SDG-4 achievement (β = 0.53). By integrating control-value theory and the job demands-resources model, this study provides a robust framework for understanding digital stress dynamics in language learning. Practically, the findings underscore the urgency of embedding digital wellness training, culturally responsive onboarding for international students, and discipline-specific AI ethics guidelines into EFL curricula to safeguard learner well-being and advance equitable, sustainable education.