AI-mediated informal digital learning of English in Kazakhstan and Uzbekistan: The roles of L2 motivational selves and learner resilience

Abstract

The rapid development of generative artificial intelligence (AI) has reshaped informal digital learning, yet most research has focused on well-resourced contexts and often treats AI-mediated learning as a standalone phenomenon. This leaves a limited understanding of how learners in underrepresented regions adopt AI-supported practices and how such practices are connected to established forms of informal learning. Addressing this gap requires examining both the developmental relationship between IDLE and AI-IDLE and the psychological factors that shape learners’ participation. This study investigates how L2 motivational selves and learner resilience are associated with learners’ participation in IDLE and AI-mediated informal learning across two Central Asian contexts. Drawing on proactive language learning theory, the study employed a cross-sectional survey design with 997 university students from Kazakhstan and Uzbekistan. A structural equation modelling (SEM) approach was used to examine the relationships among ideal and ought-to L2 selves, learner resilience, IDLE, and AI-IDLE, alongside a multigroup analysis to test cross-context variation. The findings showed that L2 motivational selves significantly predict resilience and IDLE, while resilience supports AI-mediated learning primarily through indirect pathways. IDLE emerged as the strongest predictor of AI-IDLE, suggesting a close connection between established informal digital learning practices and learners’ use of AI-supported tools. Although the overall model was broadly stable, several relationships varied across contexts, highlighting the role of sociocultural and institutional conditions in shaping AI-mediated informal learning in Central Asian contexts.

The authors' abstract, as published at the source. Learning and Motivation, 2026 · DOI ↗