Human-Centered Multi-Agent GenAI in Adaptive Immersive Learning Environments

The convergence of multi-agent Generative Artificial Intelligence (GenAI), spatial computing, and intelligent gamification is transforming research on human learning, behavioral adaptation, and digital engagement. Unlike systems centered on a single AI interaction or static immersive content, emerging environments support multiple specialist agents, including adaptive tutors, synthetic peers, and dynamic non-player characters, that interact with one another and with human users in real time. These systems can integrate behavioral telemetry, eye tracking, psychophysiological signals, and other multimodal indicators to model cognitive engagement, affective states, and user performance. However, important questions remain about how learners interpret and coordinate feedback from multiple agents, how adaptive systems can minimize cognitive load, and how dynamic difficulty adjustment can support both playability and emotional balance. Recent advances in large language models, multimodal sensing, extended reality, and affective computing provide new opportunities for responsive personalization, yet concerns persist regarding hallucination, explainability, privacy, algorithmic bias, and the transferability of findings across educational and socio-cultural contexts. A systematic investigation is therefore needed to establish rigorous theoretical models, evaluation methods, and human-centered design principles for these evolving ecosystems.This Research Topic aims to investigate the design, orchestration, and evaluation of multi-agent GenAI systems operating within adaptive immersive environments. It will examine how specialist agents can collaborate to support learning, creativity, cultural participation, and behavioral adaptation while maintaining user autonomy, trust, cognitive manageability, and emotional wellbeing. Contributions may address how multimodal data can be used to infer cognitive and affective states, how adaptive feedback and difficulty scaling influence learning and engagement, and how agent reliability can be assessed in formal and informal learning settings. The Research Topic also seeks empirical and theoretical contributions addressing transparency, fairness, accessibility, privacy, and cultural responsiveness. Particular attention will be given to practical benchmarks and evaluation methods that connect computational performance with educational, cognitive, behavioral, and social outcomes. In line with the Chief Editors’ guidance, submissions are expected to prioritize human-centered telemetry, interaction mechanics, and direct user outcomes over pure back-end or foundational system infrastructure.To gather further insights into human-centered, responsive, and equitable adaptive immersive ecosystems, we welcome articles addressing, but not limited to, the following themes:- Human-centered telemetry and interaction mechanics, including empirical evaluation of user-side telemetry, real-time feedback loops, and direct user outcomes such as learning gains, behavioral shifts, and emotional regulation in adaptive systems.- Multi-agent GenAI architectures, including agent orchestration, collaboration, role allocation, synthetic tutors, interactive personas, and cognitive or affective evaluation.- Multimodal analytics and behavioral profiling using biometrics, eye tracking, psychophysiological measures, interaction data, and learning analytics.- Adaptive gamification and dynamic difficulty adjustment tailored to diverse cognitive, motivational, and accessibility needs.- Immersive learning applications in STEM education, virtual laboratories, augmented and virtual reality classrooms, museums, cultural spaces, and digital heritage.- Cognitive ergonomics and human factors, including cognitive load, information processing, trust, usability, presence, playability, and emotional regulation.- Hallucination prevention, factuality assessment, uncertainty communication, and quality assurance for GenAI in learning environments.- Ethical, inclusive, and accessible design, including bias mitigation, privacy protection, explainability, cultural responsiveness, and support for special-needs populations.- Theoretical models and empirical benchmarks linking agent behavior, multimodal adaptation, learning outcomes, engagement, and wellbeing.- Participatory, longitudinal, and cross-cultural methods for evaluating adaptive immersive ecosystems.- Technology reports on platforms, prototypes, datasets, simulations, and deployment experiences.We welcome original research articles, systematic reviews, perspectives, and technology reports. Topic editors: Momina Shaheen, Javed Ali Khan, Dr. Faizan Ahmad Source: https://www.frontiersin.org/research-topics/84375/human-centered-multi-agent-genai-in-adaptive-immersive-learning-environments