AI vs. human touch: Comparing non-AI and AI-mediated informal digital learning of English (IDLE) and their associations with affective variables and L2 willingness to communicate.

The advancement of digital technologies has spurred a surge in research on informal digital learning of English (IDLE) in second language (L2) acquisition. However, few studies have compared the relationships between artificial intelligence (AI) and non-AI-IDLE on affective variables such as grit, motivation, self-confidence, and L2 speaking anxiety, which are associated with L2 willingness to communicate (WTC). To address this gap, this study employed an explanatory sequential mixed-methods design in the English as a foreign language (EFL) context of Iran, collecting data through 250 questionnaires and 10 semi-structured interviews. Multiple regression modeling revealed that while both digital environments correlated with the affective variables, only non-AI IDLE, self-confidence, grit, and motivation acted as significant predictors of L2 WTC. Complementing these statistical insights, thematic analysis of the qualitative findings provided critical context for this human-centric preference, suggesting that non-AI-IDLE fostered greater community engagement and situational awareness compared to AI-IDLE. These results highlight the importance of balancing AI and non-AI IDLE to support students' affective variables and enhance their L2 WTC. This research not only contributes to a deeper understanding of IDLE and L2 communication behaviors but also emphasizes the need for further investigation into the evolving role of AI in informal language learning in the rapidly changing world.