Generative artificial intelligence is beginning to open up new possibilities for more personalized, inclusive, and sustainable university learning. The article by Rubén Juárez Cádiz, published in RIED, presents the application of the PathRAG model in an adaptive learning environment supported by generative AI.
The proposal is based on the creation of dynamic educational pathways, adjusted to students’ profiles, pace, and needs, with the aim of reducing gaps in access, participation, and academic progress in hybrid and virtual contexts.
The study was developed through a quasi-experimental design in an online master’s program, integrating PathRAG into Moodle together with a generative model based on GPT-3.5-turbo. The system used a knowledge graph with thousands of nodes and semantic relationships to recommend personalized learning itineraries, generate summaries, propose guided exercises, and provide adaptive feedback.
The results show significant improvements in active participation, conceptual understanding, skills development, perceived inclusion, and overall satisfaction. These improvements were especially relevant among students with specific educational needs, learning difficulties, sensory limitations, or technological barriers.
The main value of the study lies in showing that generative AI, when integrated according to pedagogical and ethical criteria, can contribute to a more equitable education. PathRAG does not merely provide automatic content; rather, it organizes more coherent learning pathways tailored to each student, reducing cognitive overload and fostering autonomy. Nevertheless, the study also warns of important limitations: the absence of a control group, the use of a specific sample, the limited duration of the intervention, and the risks associated with algorithmic bias, the opacity of closed models, and technological dependency.
Ultimately, the article offers promising evidence of the potential of generative AI to transform adaptive learning, while emphasizing that its impact will depend on supervised, inclusive, and transparent implementation.
---
How to Cite: Juárez Cádiz, R. (2026). PathRAG application in adaptive learning with generative AI for inclusive and sustainable education. RIED-Revista Iberoamericana de Educación a Distancia, 29(1), 267–297. https://doi.org/10.5944/ried.45378
