Agentic AI and Large Language Models in Education: Perspectives from Neuroscience, Applied Linguistics, and Learning Sciences

Large Language Models (LLMs) and agentic AI are transforming education by enabling personalized feedback, interactive tutoring, adaptive support, and sustained learner–AI dialogue. Despite their rapid adoption, there is still limited understanding of how these technologies influence the cognitive, neural, and linguistic mechanisms that underlie effective learning. Existing research often emphasizes implementation, AI literacy, learning analytics, or ethical concerns, but less attention has been given to how AI affects attention, memory, motivation, self-regulation, language development, and learner agency. Addressing these questions requires an interdisciplinary approach that integrates educational neuroscience, cognitive science, applied linguistics, learning sciences, and educational technology. This Research Topic therefore focuses on mechanism-informed studies that explain how, why, and under what conditions LLMs and agentic AI improve—or potentially hinder—learning outcomes, language development, engagement, autonomy, and long-term knowledge retention.LLMs and agentic AI have the potential to provide personalized feedback, support second-language learning and English pronunciation, facilitate discourse-based learning, and improve learner engagement. Interest in these technologies continues to grow; however, their effects on the cognitive, neural, and linguistic processes underlying learning remain insufficiently understood. How do AI systems influence students’ attention, memory, motivation, self-regulation, critical thinking, and language development? Under what conditions do they promote meaningful learning rather than cognitive dependence or superficial performance? What is the teacher’s role in increasingly AI-supported classrooms? Addressing these questions requires cross-disciplinary research integrating neuroscience, cognitive science, applied linguistics, learning sciences, and educational technology. Submissions should demonstrate methodological rigor and connect specific AI features or interaction patterns with clearly defined learning mechanisms and measurable outcomes. The Research Topic aims to generate evidence-based principles for the effective, inclusive, and responsible design of AI-supported education, while contributing to SDG 4 (Quality Education) and SDG 9 (Industry, Innovation, and Infrastructure).Topics of interest include, but are not limited to:- Cognitive and neural mechanisms underlying LLM- and agentic AI-supported learning- Effects of educational AI on attention, memory, cognitive load, motivation, and knowledge retention- AI-supported metacognition, self-regulation, critical thinking, and learner autonomy- Cognitive offloading, overreliance, and dependence on generative AI- Agentic AI tutors, neuro-tutoring systems, and adaptive learning assistants- Educational neuroscience and neurophysiological evaluation of human–AI interaction- Neuroadaptive learning and educational brain–computer interfaces- LLMs in second-language acquisition and multilingual education- AI-supported pronunciation, speaking, writing, reading, and listening development- Conversational agents and dialogue-based language learning- Discourse, pragmatics, and interaction patterns in learner–AI communication- Cognitive and linguistic mechanisms of AI-generated feedback- Human–AI co-agency, co-regulation, and teacher orchestration- Multimodal AI systems informed by cognitive and linguistic evidence- Empirical and longitudinal evaluations of learning, engagement, transfer, and achievement- Equity, accessibility, transparency, and responsible AI in relation to measurable learning processes and outcomes Topic editors: Abdelkader Nasreddine Belkacem, Omar Elharrouss, Sharif Alghazo Source: https://www.frontiersin.org/research-topics/83344/agentic-ai-and-large-language-models-in-education-perspectives-from-neuroscience-applied-linguistics-and-learning-sciences