The fact that a tool is easy to use does not guarantee that its users will trust it. That is one of the most striking conclusions of a study conducted with 316 university students in Argentina, which examined what factors lead young people to place their trust in ChatGPT for their academic tasks.
The researchers drew on a classic theoretical model in the field of technology adoption and adapted it to the university context, measuring three key variables: students' optimism toward technology, their perception of the tool's usefulness, and the ease with which they use it. The underlying question was simple but significant: what makes a student trust artificial intelligence for learning, or not?
The results point in a clear direction. Technological optimism carries considerable weight: students with a positive attitude toward new technologies perceive ChatGPT as more useful and, as a result, trust it more. But the surprise comes with ease of use: even when the tool is intuitive and straightforward to operate, that does not translate directly into greater trust.
What truly matters is that students perceive a concrete benefit, that they feel ChatGPT helps them solve a real problem, write more effectively, or research more quickly. In other words, trust does not come from something being comfortable to use, but from it being genuinely useful.
The practical implications of the study are straightforward. If universities want their students to integrate artificial intelligence effectively and responsibly, simply providing access to the tool is not enough: they need to show students what it is specifically good for in their academic lives. Hands-on workshops, exercises applied to real courses, tangible examples of use in essays or research projects.
At the same time, the authors warn that trust must not become dependence: a student who trusts ChatGPT also needs to know when to question it, how to cross-check its responses, and what its limitations are. Trust in AI, yes, but with judgment.
---
How to Cite: Marimon, F., Arias Valle, M. B., Coria Augusto, C. J., & Larrea Arnau, C. M. (2025). From optimism to trust: how ChatGPT is reshaping student confidence in AI-driven learning. RIED-Revista Iberoamericana de Educación a Distancia, 28(2), 131–153. https://doi.org/10.5944/ried.28.2.43238
