Student AI surveys should ask about competence, authorship, and fairness
Updated Jun 21, 2026
Universities are getting better at asking whether students use Generative AI. They are still much worse at asking what students think AI is doing to competence, authorship, and fairness in academic work. At Student Voice AI, we see the same gap in local AI surveys and module evaluations. That is why Wendy Wenxi Hu's Studies in Higher Education paper, "University student perspectives on generative AI: reconfiguring competence, fairness, and authorship in academic work", matters. For UK institutions building on recent evidence about student experiences of GenAI across universities, it shows why AI-related student voice gets more useful when it moves beyond usage rates and into judgement, ownership, and trust.
Context and research question
Much of the current university conversation about GenAI still revolves around two questions: are students using it, and are they crossing a line when they do? Both questions matter, but they flatten important differences. A student who uses AI to test an interpretation, compare sources, or improve wording is doing something very different from a student who relies on it to avoid thinking through the task. If surveys only ask about frequency or usefulness, they miss that difference.
Hu's paper tackles that problem through interview data from 17 students across disciplines in Hong Kong. The study examines how students use GenAI in tasks such as essay writing, literature reviews, and coding, then asks how those practices are reshaping ideas of competence, authorship, and fairness. The context is not UK-based, but the practical question travels well because UK universities are already revising AI guidance, assessment policy, and local survey design around the same tensions.
Key findings
Students did not describe GenAI as a neutral efficiency tool. The abstract shows them using it selectively while deciding which tasks to delegate and which to keep under human control. The important shift is conceptual. Competence starts to look less like doing every step alone and more like exercising judgement about when, why, and how to use AI well. For UK universities, that means a simple question about whether students use GenAI tells you very little about the quality of that engagement.
Authorship was being renegotiated, not abandoned. Students used AI to support fluency or structure, but still wanted ownership of ideas and voice. That matters in higher education settings where GenAI may help students get started, phrase an argument more clearly, or work through uncertainty without wanting the final output to feel outsourced. The practical point is that authorship concerns may show up in feedback as unease about legitimacy, confidence, or loss of voice, not only as formal misconduct.
Fairness also emerged as a contextual issue rather than a simple detection problem. The abstract says students moved towards an understanding of fairness that emphasised human-AI collaboration and contextual discretion over universal detection rules. That suggests fairness concerns are tied to assessment purpose, local expectations, and the reasonableness of course-level guidance, not only to whether AI was used at all.
The paper captures the scale of that shift clearly:
"students are not simply adapting to AI but actively reshaping what academic work and learning mean"
The broader contribution is to move AI student voice away from output and towards educational purpose. The paper argues that competence, authorship, and fairness are being reworked together. For universities, that is a warning against surveys or policies that treat AI as a narrow integrity issue. Students are interpreting it through the meaning of learning itself.
Practical implications
For UK higher education teams, the first implication is to rewrite AI survey questions so they separate use from judgement. Ask not only whether students used GenAI, but whether they used it to start work, test ideas, improve wording, compare sources, or check understanding. Then ask where they felt the boundary between support and authorship became unclear. That gives teams evidence they can act on, not just a headline usage rate.
Second, universities should treat AI-related student voice as more than an academic integrity pulse check. Recent sector discussion on why student voice on AI needs more structure points in the same direction as this paper: students' views on fairness, clarity, ownership, and confidence need to be gathered explicitly if policy is going to reflect lived experience. The benefit is more targeted action on guidance, assessment design, and skills support.
Third, institutions should read AI comments as signals about support needs as well as policy risk. If students turn to GenAI because they lack confidence in writing, need a private way to test ideas, or are unsure what counts as legitimate help, the right response may sit partly in academic skills, feedback design, or induction rather than enforcement alone. That gives universities a better chance of fixing the underlying friction rather than only policing the symptom.
Finally, if universities are collecting more AI-related open comments, they need a reproducible way to analyse them. A governed workflow such as our NSS open-text analysis methodology makes it easier to separate recurring themes about competence, authorship, fairness, and policy confusion, especially when teams are comparing local interpretations with generic LLM approaches. Student Voice Analytics fits here because it helps institutions group those signals consistently across surveys and module feedback. The takeaway is clearer evidence before policy changes harden into institutional routine.
FAQ
Q: How should a university redesign AI survey questions after reading this paper?
A: Start with four clusters: what students used AI for, what they kept under human control, where authorship felt unclear, and whether local rules felt fair in practice. Then add one open-text prompt asking what felt helpful, uncomfortable, or hard to judge. That gives Student Experience and Quality teams a more usable basis for action than a simple question about whether students used GenAI.
Q: What are the methodological limits of the study?
A: This is a qualitative interview study with 17 students across disciplines in Hong Kong. Its value lies in explaining how students make sense of competence, fairness, and authorship, not in estimating sector-wide prevalence. UK universities should therefore use it as a strong design prompt for local surveys and comment analysis, not as a direct benchmark.
Q: What does this change about student voice on AI more broadly?
A: It suggests AI-related student voice should not be read as simple pro-AI or anti-AI sentiment. Comments and survey responses can carry information about judgement, ownership, fairness, confidence, and educational purpose all at once. Universities that listen for those distinctions will be better placed to write clearer policy and design more credible support.
References
[Paper Source]: Wendy Wenxi Hu "University student perspectives on generative AI: reconfiguring competence, fairness, and authorship in academic work" DOI: 10.1080/03075079.2026.2658738
Request a walkthrough
Book a free Student Voice Analytics demo
See all-comment coverage, sector benchmarks, and reporting designed for OfS quality and NSS requirements.
- All-comment coverage with HE-tuned taxonomy and sentiment.
- Versioned outputs with TEF-ready reporting.
- Benchmarks and BI-ready exports for boards and Senate.
UK-hosted · No public LLM APIs · Same-day turnaround