New research from Coursera, an online learning platform, has revealed a sharp rise in AI use and confidence among UK university students, while sentiment among university educators has declined. The findings highlight a widening sentiment gap between students and university staff as AI becomes more deeply embedded in higher education.
According to the 2025 report, 85% of UK university students believe AI is having a positive impact on higher education, a significant increase from 67% last year. In contrast, educator positivity has dropped to 69%, down from 85% last year. This represents a reversal of 2024 sentiments, when educators were more optimistic about AI’s role in higher education than their students.
Students are now far more likely to believe that AI has significantly improved the quality of higher education, with 68% saying it has done so (versus 44% in 2024). Among educators, that belief has declined from 72% to 58%.
Student positivity can be connected to their belief that AI use is improving their attainment and enabling personalised learning at scale. Four in five (80%) university students say their grades have improved since using AI, with 43% saying they have significantly improved. This is markedly higher than last year, when just over half (52%) said their grades had improved (with 8% identifying substantial improvement).
In addition, when asked to identify specific ways in which AI could have a positive impact on higher education, the most commonly cited example is promoting personalised learning (mentioned by 41% of UK respondents).
Yet concerns about AI’s potential risks are rising among university staff: while in 2024 just over a fifth (21%) did not believe AI was having, or could have, a negative impact on higher education, that figure has now fallen sharply to just 3%. Educator confidence in detecting AI-generated work has also fallen, from 42% in 2024 to 26% this year.
AI usage continues to accelerate across UK universities
The research indicates that students are increasingly integrating AI into everyday learning, with AI now used to complete 48% of their studies — double the proportion reported in 2024 (24%). Students report using AI most frequently for:
- Time management (48%)
- Writing content such as essays and assignments (45%)
- Exam revision (44%)
Use of AI for research has declined, falling from 56% in 2024 to 42% in 2025.
Educators’ perceptions of student usage have aligned more closely with student norms over the last twelve months. This year, educators estimate that 46% of students’ studies involve AI, closely aligned with what student’s report. Last year, educators believed students were using AI far more extensively than they were, estimating 43% versus the reported 24%.
Despite declining optimism about AI’s overall impact, faculty usage remains high. Sixty-eight percent say they use AI often or always in their work, consistent with 69% in 2024. However, the proportion who use AI ‘all the time’ has risen from 21% to 33%.
Educators are using AI most frequently for:
- Drafting correspondence (e.g., emails) to students (35%)
- Planning tutorials and lectures (35%)
- Marking and grading exams (34%)
However, only 27% now believe they and their colleagues have the right skills to use AI effectively, down from 37% in 2024 — suggesting growing discomfort about capability and training as AI’s development accelerates.
Acceptance of AI grows; confidence in governance and training weakens
As usage increases, acceptance of AI’s role in higher education is also rising. Both educators and students are now less likely to believe that the UK higher education system is unprepared for rapid AI adoption: 46% of educators and 47% of students say the system is unprepared, an improvement from 62% and 52% respectively.
There has also been a significant shift in attitudes towards regulation: Fifty-five percent of both educators and students now believe that the credibility of degrees will be undermined without AI regulation, down from 74% and 68% respectively in 2024. While still a majority view, the trend reflects growing acceptance that AI can be compatible with academic integrity and high pedagogical standards.
However, awareness and confidence in institutional governance have declined. Only 30% of UK educators say their university has a policy in place regarding AI usage, down from 48% last year.
Marni Baker Stein, Chief Content Officer at Coursera, said: “AI is now ubiquitous throughout higher education. Students are increasingly embracing AI to support their learning, while educators are also adopting the technology more widely, yielding productivity gains. However, many educators still require additional support to build the confidence and skills needed to use AI to its full potential and assuage concerns about implementation.
“Although acceptance of AI in higher education is growing as both students and educators shift towards practical integration, our results highlight the importance of clear institutional guidance to safeguard academic integrity and institution-wide AI upskilling initiatives to drive faculty confidence. Coursera remains committed to supporting learners and educators to use AI responsibly and effectively and to collaborating with institutions to ensure that technological progress continues to serve equitable, world-class education.”
