In an information ecosystem where an answer can appear flawless without being true, universities face a problem that runs deeper than academic misconduct or task automation. In "Truth and Truthfulness: Universities at the Crossroads of Generative AI", Faraón Llorens-Largo and Rafael Molina-Carmona argue that artificial intelligence does not make truth disappear, but it does make it more costly to identify: it multiplies plausible content, blurs the signals of reliability, and requires us to devote greater effort to verifying information that could previously be accepted with relative confidence.
The article thus shifts the debate away from the familiar question of which tools should be permitted in the classroom and towards a broader one: what must universities do to remain institutions capable of distinguishing knowledge from opinion, conjecture, and convincing simulation?
One of the paper’s main strengths is the connection it establishes between this crisis of truthfulness and university sovereignty. The authors warn that adopting external platforms, models, and services is not merely a technical decision, since it entails relinquishing some control over data, procedures, and the conditions under which knowledge is produced. In response to this dependence, they call for digital sovereignty, ethical governance, and critical literacy extending not only to students, but also to faculty members and institutional leaders.
This reflection takes concrete form in the “cube of truth, truthfulness, and trust”, a framework structured around three questions: whether the knowledge is already familiar or new to the student, whether it can be verified through formal criteria, and whether the task involves imitation or original creation. The eight resulting combinations help explain why not all academic uses of AI carry the same level of risk and why oversight must be adapted to the nature of each activity.
The article stands out for its conceptual ambition, its ability to connect philosophy, technology, pedagogy, and university policy, and its reminder that efficiency cannot become the guiding principle of higher education. Its proposal should, however, be understood as a tool for structuring reflection rather than as an empirically validated model, a limitation that the authors themselves acknowledge.
The cube inevitably simplifies complex educational realities and will need to be tested in specific contexts of teaching, research, and administration before its practical value can be established. Despite this limitation, the article advances a particularly valuable idea: in an age of automated content production, universities will not remain relevant by generating more texts or producing them more quickly, but by guaranteeing traceability, responsibility, corroboration, and sound judgement. Their distinctive task will be to continue providing well-founded reasons for trust.
---
How to Cite: Llorens-Largo, F., & Molina-Carmona, R. (2026). Truth and truthfulness: Universities at the crossroads of generative AI. RIED-Revista Iberoamericana de Educación a Distancia, 29(2), 63–80. https://doi.org/10.5944/ried.47174