Beyond the Digital Divide: Exploring Literacy, Perspectives, and Barriers to GenAI Among K12 EFL Teachers in China's Under‐Resourced Regions

ABSTRACT GenAI could either exacerbate or alleviate inequalities in under‐resourced K12 EFL education, depending largely on teachers’ ability to utilize it. However, these teachers remain understudied. To address this digital divide, this mixed‐method study, grounded in Bandura's Social Cognitive Theory, aims to provide a comprehensive understanding of these teachers’ GenAI integration by exploring their GenAI literacy, perspectives, and barriers. Latent profile analysis of quantitative data ( n = 514) in Mplus 8.7 identified three distinct profiles: lagging (14.1%), moderate (70.6%), and advanced (16.3%). Nominal logistic regression showed no significant influence of gender, age, teaching experience, or educational background on profile membership. However, teachers from primary schools, townships, and those with senior titles were more likely to belong to the lagging group. Thematic analysis of qualitative data from open‐ended questionnaires ( n = 313) in MAXQDA 2022 revealed several key insights. 21.4% of teachers had never used GenAI. The remaining teachers identified eight key benefits (e.g., creating immersive situations) and risks (e.g., ethical concerns). Four major barriers were also highlighted (e.g., insufficient training and guidelines, lack of resources). Based on these findings, we call for immediate support to ensure educational equity for these teachers.