Transforming language teacher education in the Global South

Generative artificial intelligence is reshaping language education, yet teacher training in many Global South countries remains uneven, and evidence on CELTA teacher readiness remains limited. This study examined the readiness of CELTA-certified English language educators in Saudi Arabia, Turkey, Pakistan, and Uzbekistan to use generative AI via a convergent mixed-method design. Data were collected from 120 teachers through task-based performance assessments guided by the artificial intelligence assessment scale, semi-structured interviews, and digital teaching logs. The results show strong functional AI literacy and pedagogical adaptability, alongside weaker ethical awareness related to data privacy, algorithmic bias, and responsible use. Differences across countries were more strongly associated with institutional support, infrastructure, and policy alignment than with teaching experience. The findings support the embedding of AI literacy and ethics into CELTA-oriented preparation to promote equitable, context-responsive teacher education.