The university experience has changed drastically since artificial intelligence (AI) became widely available four years ago.
Students are increasingly using AI for around-the-clock feedback, and most are required to use it as part of new assessment redesigns.
But as AI is used more in courses and changes how university works, how do students feel about it?
Our research
In 2024, we began a research project across four Australian unis – Monash University, the University of Queensland, Deakin University and the University of Technology Sydney – asking students about AI.
Undergraduate and postgraduate students were asked how frequently they use AI for study, their most common uses of AI, why they choose to use it and their concerns around the tech.
This year, we repeated the survey at the same unis, hearing from more than 10,000 students.
The survey included responses from neurodivergent students, students with disability, those who spoke languages other than English at home and students who were the first in their family to attend uni.
Students were recruited via email, posts on online university platforms, and signs with QR codes. We also ran 20 focus groups with 79 students online in 2024.
What we found
A published snapshot of our 2026 survey data shows the most common reason students use AI is to edit or improve their writing, with nearly 70% selecting this option. About 64% of students use it to generate ideas or examples, and 60% said they use AI to teach them how to do something.
Around half of students seek out information through AI and also ask it for feedback on their work.
We also know more students are using AI. When comparing the data to 2024 responses, the proportion of students who reported never using AI for their studies fell slightly, from 17% to 15%.
It mirrors findings from this year’s Australian Student Wellbeing Report, which found 84% of students have used AI tools for assignments and study, increasing 10% from the previous year.
Why students use AI
Students reported using AI for a range of different reasons.
About 55% use it to speed up their working process and also to help themselves overcome mental blocks. Nearly half of students said they use AI to make study easier in general.
Participants also reported using AI to help improve their personal wellbeing during study sessions. About 41% said the tech helps reduce stress and around 32% said they feel more confident when using AI in their work.
There are concerns
But using AI does not necessarily mean students are comfortable with it. We found many are limiting their AI use, some are trying to avoid using it, and 15-17% are avoiding it altogether in their studies.
The reasons for limiting AI use vary, but the most common explanation, selected by 67% of participants who use AI, related to concerns around AI’s accuracy.
About 62% of students said they limited or avoided AI use as it conflicts with their personal learning values and around 57% said they preferred to do the work themselves. Just under half of students expressed ethical concerns in using AI. These concerns might relate to AI presenting biased information, privacy issues, threats to academic integrity, and the environmental impacts of AI technologies.
Over the last two years, students have become less worried about breaking uni rules too. This year, 56% of participants said they avoided using AI out of a fear of being caught – decreasing 7% when compared with 2024 data.
Unis can help
Over the last four years, unis have been navigating complex changes. But their guidance on AI hasn’t always been clear to students.
Research shows students have learned more about AI on their own than they have from their unis.
Our 2024 survey and focus groups also showed widespread anxiety and confusion among students about using AI for study, partly due to confusing university academic integrity policies around AI rules.
One participant at the time asked for clearer rules:
just have some kind of concrete policy in place that you can look at and go like, ‘right, this is okay. This isn’t. Thank you.’ And even if it’s something I would personally disagree with, at least it’s set in stone.
Another student recognised the complexity of AI regulation too:
we’re in this very grey area where nobody knows what’s really happening.
Guidance around AI has improved
Our findings show there has been some progress over the last two years. Students who reported receiving sufficient guidance from unis rose from 36% in 2024 to 47% in 2026.
This might be because unis are setting more consistent AI guidelines across their institutions, rather than putting the responsibility on educators to set rules around AI use for each subject. This is happening alongside changes to assessment that incorporate use of AI.
But our recent survey suggests there is more to do to help students feel confident they will not break rules.
Students’ perspectives matter
Of all the concerns about AI in higher education, perhaps one of the most important questions is whether unis are doing enough to listen to students about their feelings towards AI and how they use it.
When unis don’t engage students in these conversations, they miss opportunities to understand how AI is influencing learning. More importantly, they miss the opportunity to involve students in shaping how AI is used in education.
Kelly Matthews has received funding from the Office of Learning and Teaching (now ceased) and the Australian Department of Education as commissioned work for the Higher Education Standards Panel. The research mentioned in this article has been funded by the Deputy Vice-Chancellor Education Portfolios at Monash University, University of Queensland, Deakin University and University Technology Sydney
Tim Fawns does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.
