ABSTRACT Background Amid Bangladesh's national efforts to enhance digital inclusion, rural university students continue to face persistent infrastructural and educational inequalities that limit their access to quality English learning opportunities. While informal digital environments increasingly support learners' out‐of‐class English development, little is known about what factors predict rural learners' participation in such practices, especially as artificial intelligence (AI) tools reshape the digital learning ecology. Addressing this gap is critical for ensuring that digital learning benefits learners in resource‐limited contexts. Objectives Guided by the proactive language learning theory, this study aims to identify the sociocontextual, motivational and affective factors that predict rural Bangladeshi students' involvement with informal digital learning of English (IDLE) and its emerging form, AI‐mediated IDLE (AI‐IDLE). Methods We collected data from 508 undergraduate students from rural Bangladesh using an online survey. Hierarchical regression analyses were conducted to examine the extent to which sociobiographical, sociotechnical, motivational and affective variables predict IDLE and AI‐IDLE. Results and Conclusions Self‐efficacy, enjoyment and the ideal L2 self significantly predicted IDLE, while AI‐IDLE was positively predicted by IDLE and the ought‐to L2 self but negatively predicted by the ideal L2 self, with university type also showing a significant effect. These findings highlight that learners' affective and motivational dispositions, rather than demographic or sociotechnical factors, are central to shaping informal English learning with technology in underrepresented contexts. The study also underscores that prior informal learning experience provides a foundation for AI tool adoption for out‐of‐class learning purposes and advances an inclusive understanding of how Global South learners engage proactively in evolving digital learning ecologies.
Liu et al. (Fri,) studied this question.