What happens when a ChatGPT-based assistant stops being an external tool and becomes part of the virtual campus itself? Miguel-Ángel Cabeza-Rodríguez examines this experience through the integration of a ChatGPT 3.5-based assistant into 21 courses at an online university, spanning a wide range of degree programs and fields of knowledge.
The study gathers the views of 391 students and examines four dimensions of their experience: the assistant’s efficiency, its impact on learning, its contribution to skill development, and technical and accessibility aspects.
The results reveal a largely positive perception. Students particularly value the assistant’s ability to clarify concepts, answer questions, and support more autonomous learning, to the point that many participants would like to see it extended to other courses.
The analysis also shows that satisfaction depends above all on whether the assistant is genuinely effective: its ability to provide useful answers matters more than purely technical considerations. At the same time, the experience also exposes the limitations of these tools. Among the main criticisms are answers that lack precision or sufficient depth, the absence of certain functionalities, and the need to provide students with better guidance on how to use AI effectively.
The question, therefore, is not simply whether universities should introduce AI assistants, but how to do so in ways that create genuine educational value. Students tend to see them primarily as a complement that can provide immediate support and encourage autonomous learning, rather than as a substitute for teachers or critical thinking.
The experience therefore points towards a model in which the integration of artificial intelligence must be accompanied by sound pedagogical design, sufficiently reliable responses, guidance on how to use the tool, and proper alignment with the specific needs of each course. The potential is clearly there, but its educational value will depend less on simply making a chatbot available than on turning it into a genuine assistant for learning.
---
How to Cite: Cabeza-Rodríguez, M.- Ángel. (2025). ChatGPT assistants in online higher education and student satisfaction: a case study. RIED-Revista Iberoamericana de Educación a Distancia, 28(2), 9–38. https://doi.org/10.5944/ried.28.2.43552
