A recent study of *1,097 students across two Swedish universities* offers a useful perspective on how students are actually using Generative AI in higher education.
The findings are interesting.
Students generally see GenAI as a *supporting tool, not a replacement for teachers or traditional learning*.
Common uses include:
– Finding and understanding information – Improving and analysing text – Discussing ideas – Tutoring and getting explanations
At the same time, students showed only *moderate trust in AI-generated answers*. That matters. They appear to recognise that AI can be helpful, but they also understand that its answers need to be checked.
Perhaps the most important point is how GenAI is used.
If a student asks AI to simply produce an answer, the learning opportunity may be limited. But if the student uses it to question an idea, compare explanations, identify gaps, or reflect on an answer, it can become part of a deeper learning process.
There is a small but important caution here: the study measured students’ reported perceptions and use of GenAI. It did *not directly measure whether their learning actually improved*.
For educators, this suggests that the question may not be *“Should students use AI?”*
A better question could be:
*“How do we teach students to use AI in ways that strengthen their own thinking?”*
That shift in thinking may be more useful than simply trying to restrict the technology.
*Source:* link.springer.com/article/10.1007/s10639-026-14100-2