Lecturers and students see AI use very differently, study finds
Academics overestimate how much students use generative AI for their studies, highlighting the need for a better understanding of how the technology is being used
Lecturers and students have significantly different perceptions of how much generative artificial intelligence (AI) is being used in university studies, according to new research from The University of 麻豆破解版 and Spain鈥檚 Universidad del Pais Vasco/Euskal Herriko Unibertsitatea.
The study found that academic staff perceived students to be using AI tools such as ChatGPT for their studies much more extensively than students reported themselves. Researchers say the findings highlight a gap in understanding between staff and students, at a time when universities are grappling with how best to respond to the technology.
- Academic staff rated their perception of students鈥� AI use at 5.23 out of seven, compared with an average of 3.11 among students reporting their own use for academic purposes.
- Students were more positive than staff about the potential benefits of generative AI, including its ability to improve productivity, support learning and reduce workloads.
- The most prominent uses reported by students involved producing written text, finding answers to assignments, and translating or improving grammar.
- Staff varied considerably in their experience and expertise with AI, making it difficult for some to judge how students were using the technology.
- Researchers argue that universities need to move beyond treating AI primarily as a technical challenge or a threat to academic integrity, and develop a better understanding of how students and staff engage with it.
A gap in understanding
The research examined how academic staff and students perceived their own use of generative AI tools, as well as their views of how the technology was being used in higher education.
The study surveyed 45 academic staff and 86 students in a School covering social science disciplines at a UK research-intensive university. The questionnaires were conducted between April and June 2024.
When asked about students鈥� use of AI for academic purposes, staff gave an average rating of 5.23 on a seven-point scale. Students鈥� own responses produced an average of 3.11.
The researchers caution that students may have underreported their use of AI, but say the difference nevertheless points to the limited information staff may have about how students use these tools in their studies.
AI use can be difficult for lecturers to identify, particularly when they see only the final version of an assignment rather than the process through which it was produced. Students may also use AI independently of their interactions with teaching staff, leaving academics with an incomplete picture of their working practices.
Different attitudes towards AI
The research also found that students were generally more positive than staff about the potential benefits of generative AI.
Students rated the technology more highly in terms of its likely contribution to academic performance, including helping them learn about their subjects, increasing productivity and reducing workloads. They also felt more strongly than staff that appropriate conditions and support were in place to help them use AI tools.
Among the uses identified in the survey, students reported using generative AI to find answers to questions and assignments, generate written work, and translate or improve the grammar of their writing.
The researchers suggest that the immediate appeal of AI tools may lie in their ability to produce and improve text. However, they argue that universities need to understand the range of ways students use the technology, rather than assume that its role in academic work is straightforward.
The study also found considerable variation in staff members鈥� familiarity with AI. While 63 per cent of academic staff respondents had used at least one generative AI tool in the six months before the survey, some had limited practical experience of the technology. Only 14 per cent had a paid subscription to an AI tool.
The authors argue that the challenges posed by AI in higher education cannot be resolved simply by providing technical training or improving students鈥� AI literacy. Instead, universities need to consider how staff and students understand the technology, what they believe it can do, and how those beliefs shape its use.
Rethinking the response to AI
The researchers say that assuming students are using AI primarily to circumvent assessment rules risks creating a relationship based on distrust. While acknowledging that AI can be used in ways that undermine academic integrity, they argue that universities need a more nuanced understanding of students鈥� practices.
They suggest that teaching and assessment should evolve in response to the capabilities of generative AI, with staff and students better supported to explore its appropriate educational uses.
The authors also stress that the research offers a snapshot of a particular institutional setting at a specific point in time. As AI tools continue to develop rapidly, perceptions and patterns of use may change, and further research with larger samples is needed.
What the experts say
鈥淭here is a tendency for the debate about AI in universities to focus on the risks of cheating or the promise of transforming education, but the reality is more complex,鈥� said Dr Mark Carrigan, Senior Lecturer in Education at The University of 麻豆破解版 and one of the study鈥檚 authors who has also recently on this topic. 鈥淥ur findings suggest that staff and students have quite different perceptions of how AI is being used, making it difficult for universities to develop effective responses based on assumptions rather than a clear understanding of students鈥� experiences.鈥�
We need to move beyond a climate of suspicion and focus on how these tools can be used appropriately to support learning, while rethinking teaching and assessment for an AI-enabled world.
Publication details
This research was published in: Frontiers in Education
Title: Still emerging: understanding Generative AI use in Higher Education
DOI: 10.3389/feduc.2026.1885253
URL: