Engagement in AI-Mediated Informal Digital Learning of English: An Adaptive Intelligence Perspective on Motivational, Psychological, and Technological Resources

Using adaptive intelligence as a theoretical lens, the study investigates how L2 motivational selves are associated with AI-IDLE engagement through psychological capital (PsyCap) and L2 grit and whether these indirect associations vary with AI literacy among 1025 Chinese university EFL learners. The proposed model was tested using structural equation modeling (SEM). For the ideal L2 self, associations with engagement emerged both directly and indirectly through PsyCap and L2 grit, whereas the ought-to L2 self showed an indirect association through PsyCap and no direct association with engagement. Higher AI literacy corresponded to stronger associations of both PsyCap and L2 grit with engagement. The two motivational self-guides were therefore associated with engagement through different psychological pathways, with AI literacy related to variation in the strength of these pathways. From an adaptive intelligence perspective, this conditional pattern indicates that the relevance of PsyCap and L2 grit to AI-IDLE engagement varies with learners’ AI literacy. These findings extend the application of adaptive intelligence to informal human–AI learning.