For every 100 euros invested in an intelligent learning platform, how much of it actually serves its purpose if students do not trust it?
No Tunisian study quantifies this loss, but a survey conducted among 321 students provides the key. The most capable technology is useless if it does not first win the trust of those who use it. Before even asking what AI can do, universities should ask whether their students grant it the right to do so.
The moment is not trivial. The country is about to deploy ELM (E-learning Mentor), a national AI-powered learning mentoring program, in which all institutions have pledged to participate for the 2024–2025 academic year. Asking students just before this turning point is to photograph a mindset before habit erases it.
This snapshot is the work of a study published recently. Between March and May 2024, 321 valid responses were collected across four institutions: ISG Tunis (36.1%), FSEG Tunis (26.2%), FSEG Sousse (22.1%), and FSJEG Jendouba (15.6%). The sample comprises 55.5% men and 44.5% women, 59.5% bachelor’s students and 40.5% master’s students.
Ease of use weighs twice as much as usefulness
First insight, counterintuitive: what breeds trust is not primarily the promise of effectiveness. The perceived usefulness of these tools positively influences students’ trust, with a coefficient of 0.122. The ease of use, however, carries more than double that weight, at 0.257. In other words, a student trusts a tool more that can be tamed without effort than one that promises a lot but confuses.
For decision-makers, the message is clear: an intelligent but austere platform risks generating mistrust, while a simple and intuitive interface builds a stock of trust from the first clicks.
Trust, the pivot of the whole dynamic
Once established, trust becomes the true engine of the system. It alone accounts for a substantial share of student engagement (0.352) and of their satisfaction (0.416). Satisfaction, in turn, drives engagement (0.383), drawing a virtuous circle: the student who trusts the tool is more content, and the more content they are, the more they invest themselves.
This mechanism is described as a chain whose first link is trust. If that link gives way, everything else shakes, regardless of the algorithm’s sophistication.
Ethics, the amplifier no one expected
