Cognitive and Affective Co-Construction in Learner–AI Interaction: Psychological Dynamics of Learning in AI-Mediated Instructional Contexts

Artificial intelligence, including large language models, conversational agents, and adaptive learning systems, is reshaping how learners think, feel, and learn. Learning environments are shifting from teacher-centered or learner-centered spaces toward dynamically co-constructed ones, where learners and AI jointly take part in meaning-making. Existing research has mostly examined the instructional affordances and outcomes of AI tools, leaving the underlying psychological dynamics underexplored: how learners and AI systems co-construct knowledge, emotion, and identity. Co-constructivism, rooted in Vygotskian sociocultural theory and Piagetian constructivism, holds that knowledge emerges through dialogic, social, and emotional engagement among participants. Generative AI calls for a reconceptualization of this framework, since it is no longer a passive tool but an active interlocutor contributing to the joint construction of meaning. This raises the question of how attention, motivation, emotion, and identity are renegotiated when AI participates directly in learning.This Research Topic aims to reconceptualize learner–AI interaction as a process of cognitive and affective co-construction, in which meaning, knowledge, and emotion are jointly produced through ongoing negotiation between learners and AI systems, with teachers mediating this process. The focus is on the psychological mechanisms through which AI shapes, and is shaped by, learners: attention, cognitive load, co-regulated and socially shared cognition, motivation, self-efficacy, emotion, trust, identity, and agency. Submissions examining how AI-mediated environments affect learners' co-regulated cognition, emotional experience (including anxiety, enjoyment, and technostress), willingness to communicate, rapport with AI, beliefs about learning with AI, intercultural and identity negotiation, ethical sensibilities, and the redistribution of agency between human and machine participants are invited. Dialogue- and feedback-intensive settings, such as second-language, writing, and intercultural learning, along with adaptive and recommender systems that model learner interests and preferences, offer a revealing window onto these dynamics and are of particular interest when linked to psychological process.Relevant themes include, but are not limited to:• Co-construction of knowledge in learner–AI dialogue;• Attention, cognitive load, and collaborative cognition;• Motivation, self-efficacy, anxiety, and well-being;• Feedback literacy and uptake of AI-generated feedback;• Learner identity, agency, and beliefs about learning; trust and rapport with AI;• Intercultural negotiation in AI-mediated communication;• Ethical engagement with AI as a learning partner;• Personalization and its effects on motivation and agency;• Teacher mediation as it shapes learner cognition and emotion.Studies may be set in classroom or informal settings and may use experimental, longitudinal, ethnographic, discourse-analytic, mixed-methods, or data-driven approaches, including interaction analysis, learning analytics, structural equation modelling, eye-tracking, and behavioral trace data. Empirical studies, systematic reviews, meta-analyses, and conceptual papers are welcome. Submissions focused mainly on tool development, platform evaluation, language proficiency gains, or teacher training, without addressing learners' psychological processes, fall outside scope. Topic editors: Azzeddine Boudouaia, Rustam Shadiev, Ankhtuya Ochirbat Source: https://www.frontiersin.org/research-topics/85974/cognitive-and-affective-co-construction-in-learner-ai-interaction-psychological-dynamics-of-learning-in-ai-mediated-instructional-contexts