Tabriz University of Medical Sciences Research and Development in Medical Education 2322-2719 15 1 2026 01 01 Effectiveness of an AI Assistant Versus Acceptance and Commitment Therapy in Enhancing Academic Engagement and Social Interaction Among Bilingual Students 33406 33406 10.34172/rdme.33406 EN Khaled Loveimi https://orcid.org/0009-0008-7794-9608 Sasan Bavi https://orcid.org/0009-0000-4696-1526 Reza Johari Fard https://orcid.org/0000-0003-0505-0501 Journal Article 10.34172/rdme.33406 2025 11 22 2026 06 25 Introduction: Bilingual students frequently encounter unique complexities within educational environments, which can pose distinct challenges to their academic engagement and social interaction. This study aimed to comparatively evaluate the effectiveness of an Artificial Intelligence (AI) Assistant and Acceptance and Commitment Therapy (ACT) in fostering academic engagement and enhancing social interaction among seventh-grade Persian-Arabic bilingual male students from Ahvaz suburban areas. Methods: Utilizing a quasi-experimental design featuring a pre-test, post-test, and one-month follow-up, alongside a control group, 105 students were recruited via multi-stage cluster sampling. Participants were subsequently randomized into three 35-member groups: AI Assistant, ACT, and control. Over eight weeks, experimental groups engaged in weekly 90-minute sessions, either interacting with the ‘Monica’ AI Assistant or participating in structured ACT therapy. Academic engagement and social interaction were assessed using standard questionnaires. Data analysis, employing Repeated Measures ANOVA with Bonferroni corrections, revealed significant main effects for time and group, as well as significant time-by-group interaction effects for both outcomes. Results: Both interventions were associated with substantial improvements in academic engagement and social interaction compared to the control group immediately post-intervention, with these gains largely maintained at the one-month follow-up. The AI Assistant group yielded significantly higher scores than the ACT group (P<0.05), although the quasi-experimental design warrants cautious interpretation of the comparative advantage. Conclusion: These findings highlight the significant potential of technology-supported learning and psychological flexibility interventions in cultivating essential non-cognitive skills within bilingual student populations and hold promising educational significance for supporting diverse learners in resource-limited settings.