Objective: This review examined recent studies (2024–2025) to identify new conceptualizations of learning in the AI era, characterized by hybrid human-system collaboration, real-time personalization using multimodal data, interpretability, and a shift from teacher-centered to learner- and technology-driven approaches.
Method: Nine peer-reviewed studies from 2024–2025 were analyzed thematically, sourced from databases including Scopus, Web of Science, PubMed, Google Scholar, and ResearchGate using keywords like "learning definition", "conceptualization of learning", "AI in education" and "hybrid intelligence."
Result: The findings reveal that learning in the AI era is conceptualized as a dynamic, adaptive, and multidimensional process integrating cognitive, emotional, social, and interactive elements. Emerging theoretical frameworks include hybrid intelligence emphasizing human-system co-evolution, the integration of cognitive load theory with AI and educational neuroscience, and generalized online learning models prioritizing interpretability and continuous autonomous learning. Key novel features encompass multimodal learning leveraging neurophysiological tools, real-time adaptive pathways, human-machine collaborative processes, incorporation of behavioral and interactional data alongside neurophysiological indicators, and a strong emphasis on transparency and interpretability
Conclusions: The core transformation is a paradigm shift toward learner- and technology-centered models, with social interaction gaining prominence in AI-mediated settings. While AI enables personalized feedback and dynamic adaptation, its limitations in context and causal reasoning underscore the necessity of human-AI synergy. Practical implications highlight the need for educator training in new technologies and careful attention to ethics and equitable access. These insights derive from a constrained sample of nine studies and may not encompass the entire scope of research on learning from 2024–2025.