Individual Differences in Language Education: An International Journal

The integration of artificial intelligence (AI) into task-based language teaching (TBLT) promises personalized learning at scale. Yet without grounding in the rich tradition of individual differences (IDs) research, AI risks reducing learners to simplified variables and substituting algorithmic accommodation for genuine cognitive engagement. This paper builds a bridge between established ID-TBLT research and emerging AI design. We critically synthesize how cognitive IDs (working memory, attentional control, language aptitude) and affective IDs (motivation, anxiety, enjoyment, boredom, and flow) interact with task complexity to shape L2 performance—acknowledging both established findings and persistent debates. From this synthesis, we articulate design challenges for AI-mediated TBLT, distinguishing what is technically feasible from what requires further research, and acknowledging inherent tensions (scaffolding vs. substitution, personalization vs. equity, detection vs. privacy, flow vs. instructional efficiency). We then extend this framework through Bui's (2026) longitudinal study of learner perceptions in AI-assisted speaking tasks, using empirical findings on learner adaptation and "prompt literacy" to illustrate and refine our design challenges. The result is a dual contribution: a systematic translation of ID-TBLT research into AI design challenges, and a refined, six-frontier research agenda for developing intelligent, adaptive TBLT systems that are cognitively grounded, affectively attuned, and ethically responsible.