What Queen Mary Students Are Telling Us About Generative AI

27 August 2026

Generative AI (GenAI) is outpacing higher education policies and is already shaping how students learn, communicate, seek support, and prepare for employment. 

 These were the key messages echoed by Queen Mary students at this year's Student Experience Forum, where discussions highlighted the growing influence of GenAI across multiple aspects of student life. 

Each year, the Forum focuses on a significant issue affecting higher education, gathers evidence from Queen Mary students and brings colleagues together to discuss what that evidence means for practice. For 2026, we explored students’ academic and non-academic use of generative AI. Our findings are based on an all-student survey completed by 508 respondents, alongside 15 in-depth qualitative interviews. These included targeted conversations with students working in English as an additional language, students with disabilities and students undertaking substantial paid work alongside their studies. While the survey is not fully representative of the entire student body, it offers valuable insight into the varied ways students are navigating this rapidly changing area. 

Our first finding is that GenAI is already embedded in student life. Over 80% of students reported using AI for some purpose. Among users, 42.9% reported using it academically at least weekly, while 25.1% reported daily use. Yet the dominant picture was not of students outsourcing whole assignments. The most common uses were understanding difficult concepts and researching or summarising articles. Students described using AI to make complex material more accessible, manage heavy reading loads and test their understanding. 

Discussions about AI in higher education are often framed primarily through academic misconduct. While integrity is essential, that lens alone misses much of what students are actually doing – and why. Students are using GenAI as a study aid, a language tool, an administrative assistant and a careers resource. The educational challenge is therefore not only to prevent inappropriate use, but to help students distinguish support for learning from substitution of learning. 

Another strong theme was scepticism. Students were not adopting GenAI uncritically. Non-users frequently cited ethical concerns and worries about misinformation, while interviewees raised questions about privacy, environmental impact, fairness and the possible erosion of reading, writing and critical-thinking skills. Attitudes also varied by discipline. Some humanities students saw the struggle involved in interpretation and writing as central to their education, while students in computing and engineering were more likely to view AI as part of their emerging professional practice. However, principled non-use should not be dismissed as resistance to innovation as it can reflect a thoughtful position on academic identity and the purpose of learning. 

The clearest area for educators to act in is developing guidance around the use of GenAI. More than a quarter of respondents said they had received no university guidance on academic use of GenAI, and a further 12.4% were unsure. 70% had avoided using it because they were uncertain about the rules. Students described a patchwork of expectations where one lecturer may prohibit AI, another may encourage it, while others may not mention it at all. Students overwhelmingly expressed wanting practical help with prompting, evaluating outputs, acknowledging AI use and developing discipline-specific AI skills.

The findings also revealed important questions of equity and inclusion. A quarter of GenAI users were paying for premium tools, in order to bypass usage limits and gain access to a more efficient GenAI model. Students felt that differences between free and paid versions could create an uneven academic playing field.  

Students reported using several accounts at once in order to work around usage limits. However, this kind of workaround requires time, confidence, technical knowledge and awareness of how different tools operate. In this sense, even navigating the free tiers effectively can become a form of advantage. Students with more digital literacy, more time to experiment or more familiarity with AI may be able to extract more value from free tools. 

For students working in English as an additional language, GenAI helps translate ideas, refine tone and participate more confidently in seminars. Used carefully, this can support expression rather than replace thought. However, a key challenge is to recognise language support as distinct from outsourcing academic work, while helping students continue to develop independent communication skills. 

Some findings extended well beyond the classroom. Students used GenAI for health questions, life administration, emotional reassurance and careers planning. Most strikingly, 34.2% said they were more likely to seek wellbeing support from GenAI than from university services. Students valued its immediacy, privacy, non-judgemental tone and ability to respond in their preferred language. This does not make AI a substitute for professional or crisis support. It does, however, tell us something important about the barriers students experience when seeking help from support services, and where signposting, cultural accessibility and timely provision may need to improve. 

So, what should educators do next? Students need clear and consistently communicated expectations, including the continued development of the RAG approach to assessment. They also need discipline-specific examples that show appropriate, inappropriate and ambiguous uses of AI in real tasks. AI literacy should include critical evaluation, privacy, consent, bias, environmental impact and the ability to decide not to use a tool. Where AI is becoming part of learning and employability, students should also have equitable access to safe, institutionally supported tools. 

Queen Mary has already begun this work through its student-facing online AI literacy course, assessment guidance and the Centre for Excellence in AI in Education. The Forum’s findings underline the importance of continuing it collaboratively, with students, educators and professional services colleagues shaping the next steps together. 

The research findings reinforced the belief that student voice must be central to institutional decisions about AI. The question is no longer whether GenAI belongs in student life; it is already there. The question is what kind of relationship with it higher education wants to cultivate. 

Without clear guidance, equitable access and opportunities to develop critical judgement, GenAI risks becoming another form of hidden advantage, benefiting students who can afford better tools, understand unwritten expectations or feel more confident experimenting. A thoughtful institutional response should not simply permit or prohibit its use. It should help every student understand what these tools can offer, where they fall short, and when learning requires something only human effort, expertise and support can provide. 

To request a copy of the full Student Experience Forum findings, please contact us at su-representation@qmul.ac.uk. 

Vandya Widyalankara: Student Voice and Feedback Assistant, QMSU

Marianne Melsen: Head of Student Voice and Insights (QMSU)