The DOI matches its Crossref record; 33 of the 34 references it cites were found in a registry or an index; 10 of the 34 sources its claims cite were checked for support.
Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education
Open accessAbstract
The rapid proliferation of Generative Artificial Intelligence (GAI) tools such as ChatGPT is transforming how students learn, create and engage with knowledge. Yet, little is known about how students in Higher Education perceive and use GAI, or how their perceptions differ across demographic groups. This study addresses this gap by examining student subjective perceptions on GAI through a questionnaire of 132 undergraduate and postgraduate students from six UK Universities, conducted in early 2024. We analyse attitudes towards usage, plagiarism, reliability and the perceived impact of GAI on the learning process, focussing on differences by gender and domicile status (UK vs. non-UK students). Our findings reveal a paradox: GAI is widely used but barely trusted. Particularly, we show that students are increasingly eager to adopt GAI for completion of coursework, assignments and other non-academic tasks, but fewer than 25% regard its outputs as reliable. Furthermore, we find differences in perceptions across student groups, with international and male students reporting higher usage and more permissive attitudes towards AI-assisted learning. These results suggest that (i) GAI adoption is driven by its usefulness and accessibility rather than by trust in the outputs, and (ii) there is a need for inclusive teaching approaches that help all students, regardless of their background, engage confidently and responsibly with GAI in their learning.
Overview
CoChat Summary
TL;DR
This study surveys 132 UK undergraduate and postgraduate students and finds a paradox: generative AI is widely used for academic and non-academic tasks but barely trusted. Usage and trust vary by gender and domicile status, with international and male students reporting higher use and more permissive attitudes.
Key findings
- Fewer than 25% of respondents across all demographic groups regard GAI outputs as reliable; UK women express the lowest trust, with only around 15% agreeing GAI outputs are reliable, while non-UK men have the largest share at approximately 25%.
- Non-UK men are the most likely to report past GAI use in coursework, assignments and university tasks, with statistically significant differences compared with UK men and UK women; across all demographics, anticipated future use is higher than past use.
- UK students are nearly unanimous in classifying GAI-generated assignment submission as academic misconduct, while non-UK students are significantly more mixed and more likely to consider it acceptable; UK students also tend to classify some institutionally acceptable uses as misconduct.
- Most students agree GAI could revolutionise education, but approximately 30% worry it could hinder learning by reducing critical engagement; UK women are the most sceptical about GAI’s role in their education.
Methods
- An online questionnaire with five parts covered demographics, usage patterns, reliability, plagiarism, and impact on learning; closed questions used Likert-type scales, with optional open-ended fields that were not analysed because few participants commented.
- A total of 132 undergraduate and postgraduate taught students from six UK universities participated in early 2024, recruited via convenience sampling; 62 identified as women (47.0%), 69 as men (52.3%), and 1 as non-binary (0.7%), with the non-binary category excluded from further analysis.
- Responses were stratified by gender (man, woman) and domicile status (UK vs. non-UK), presented through bar plots, and differences across strata were assessed with the Mann-Whitney U test.
Limitations
- The paper states that reporting of GAI use and perceptions may be affected by social desirability bias, especially for past usage, and recall bias, potentially skewing accuracy and reliability.
- The paper states that convenience sampling was used, the sample came from six UK universities, the non-binary category was excluded due to small numbers, and qualitative open-ended responses were not included in the analysis.
- Evident from the text: the sample comprised mostly students in quantitative and computational disciplines, so findings likely reflect students familiar with data-driven and technical approaches and may not generalise to other disciplines or institution types.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
9 Valid · 1 Not valid · 13 Not checked · 63 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“While many students value GAI as a complement to their learning, a substantial proportion consider its use in exams to constitute cheating (Stöhr, Ou, and Malmström 2024).”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > 1.3. Diverse student perceptions of generative AI
What the sources say
- Supportedcites Stöhr, Ou, and Malmström 2024 source
The Discussion (p. 8) reports that more than sixty percent of students believed using chatbots during examinations is cheating. The Descriptive analysis (p. 5) likewise reports that 61.9% viewed using chatbots to complete assignments and exams as cheating. (LLM, conf=0.94)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Research indicates that ChatGPT exhibits better performance in specific tasks (Sallam and Al-Salahat 2023), such as those that require general knowledge to be completed (Ray 2023).”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > 3.2. Frequent use despite concerns over reliability
What the sources say
- Supportedcites Sallam and Al-Salahat 2023 source
Section 3.2 reports that ChatGPT 3.5 performed best on Remember-domain questions (88.5% correct), with lower scores in the other cognitive domains. The difference was not statistically significant. The general-knowledge example is attributed to Ray (2023), not to this source. (LLM, conf=0.87)
- Not checked — no full text of the sourcecites Ray 2023 source
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“with international and male students reporting higher usage and more permissive attitudes towards AI-assisted learning.”
- Also stated in: § 3.1. Non-UK men report higher tendency of past and prospective GAI use · plain
“Within each of those groups, men report a higher frequency of past and prospective future use than women.”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > Abstract
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 3.1. Non-UK men report higher tendency of past and prospective GAI use · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“with international and male students reporting higher usage and more permissive attitudes towards AI-assisted learning.”
- Also stated in: § 3.3. Perceptions on GAI as a way of plagiarism · plain
“Non-UK students generally hold more liberal views regarding what constitutes academic misconduct compared to their UK counterparts.”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > Abstract
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 3.3. Perceptions on GAI as a way of plagiarism · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“with international and male students reporting higher usage and more permissive attitudes towards AI-assisted learning.”
- Also stated in: § 3.1. Non-UK men report higher tendency of past and prospective GAI use · hedged
“Broadly speaking, non-UK students tend to report a higher frequency of past and prospective future use than UK students.”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > Abstract
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 3.1. Non-UK men report higher tendency of past and prospective GAI use · hedged
43 more claims
Create a free account to see all 86 claims and to verify the 1 reference still open.
- Borderline: one of the tests for a substantial claim is unclear. 1
- It may state two separate claims as one. 28
- Probably not a substantial claim. 17
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > 3.1. Non-UK men report higher tendency of past and prospective GAI use
“UK men report higher prospective use of GAI than UK women, and the difference by gender is statistically significant.”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > 4. Discussion and conclusions
“Table A4.Q: … other tasks that are not University coursework or assignments?UK men UK women non-UK men non-UK women UK men 1.00 0.98 0.03 0.57”
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > Additional information
“UK women 0.51 1.00 , 0.01 0.02”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > Additional information
“non-UK women 0.56 0.21 0.13 1.00”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > Additional information
“non-UK men 0.12 0.74 1.00 0.38”
§ Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education > Additional information
“non-UK women 0.03 0.30 0.58 1.00”
Citations
Cites
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326
Developing and validating trust measures for e-commerce: an integrative typology
McKnight, D. H., Choudhury, V., & Kacmar, C. (2002) · Information Systems Research · doi:10.1287/isre.13.3.334.81 · Cited by 5,550
Kung, T. H., Cheatham, M., Medenilla, A., Sillos, C., De Leon, L., Elepaño, C., Madriaga, M., Aggabao, R., Diaz-Candido, G., Maningo, J., et al. (2023) · PLOS Digital Health · doi:10.1371/journal.pdig.0000198 · Cited by 4,115
Ray, P. P. (2023) · Internet of Things and Cyber-Physical Systems · doi:10.1016/j.iotcps.2023.04.003 · Cited by 2,463
Technology acceptance model: a literature review from 1986 to 2013
Marangunić, N., & Granić, A. (2015) · Universal Access in the Information Society · doi:10.1007/s10209-014-0348-1 · Cited by 2,458
Population Research: Convenience Sampling Strategies
Stratton, S. J. (2021) · Prehospital and Disaster Medicine · doi:10.1017/s1049023x21000649 · Cited by 1,220
Technology acceptance model in educational context: A systematic literature review
Granić, A., & Marangunić, N. (2019) · British Journal of Educational Technology · doi:10.1111/bjet.12864 · Cited by 1,204
ChatGPT for Education and Research: Opportunities, Threats, and Strategies
Rahman, M. M., & Watanobe, Y. (2023) · Applied Sciences · doi:10.3390/app13095783 · Cited by 1,080
Revolutionizing education with AI: Exploring the transformative potential of ChatGPT
Adiguzel, T., Kaya, M. H., & Cansu, F. K. (2023) · Contemporary Educational Technology · doi:10.30935/cedtech/13152 · Cited by 975
Individual and Contextual Influences on Academic Dishonesty: A Multicampus Investigation
McCabe, D. L., & Trevino, L. K. (1997) · Research in Higher Education · doi:10.1023/a:1024954224675 · Cited by 830
Research works printed but not linked: 2
- 10.1111/bjet.v50.5 · doi:10.1111/bjet.v50.5
- Jospeh, O. U., I. M. Arikpo, O. S. Victor, N. Chidirim, A. P. Mbua, U. M. Ify, and O. B. Diwa. 2024. “Artificial Intelligence (AI) in Academic Research. A Multi-group Analysis of Students' Awareness and Perceptions Using Gender and Programme Type.” Journal of Applied Learning and Teaching 7 (1).
Cited by
Genialt eller galt? Studentperspektiver på KI i høyere utdanning ved UiT, 2023–2025
Maxwell, K. (2026) · Septentrio Reports · doi:10.7557/7.8672
Kajan, K., Shi, W. Y., & Wanatowski, D. (2026) · Education Sciences · doi:10.3390/educsci16091358
珂, 朱., Ma, Y., Zhang, J., & Liu, J. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1940317
Metadata
- Authors
- Cabrera, C., & Neville, R.
- Affiliations
- University of Liverpool
- University College London (GB)
- ORCID
- Neville, Ruth
- Venue
- Perspectives: Policy and Practice in Higher Education
- Publisher
- Informa UK Limited
- Published
- 2025-12-16
- Pages
- 1-14
- ISSN
- 1360-3108 (ISSN-L)
- 1460-7018
- OpenAlex
- W4417410978
- Cited by
- 4
- References
- 32
- Citations by year
- 2026: 4
- FWCI
- 4.16
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Perspectives Policy and Practice in Higher Education (publishedVersion, cc-by)
- Funders
- University College London
- Newcastle University
- University of Glasgow
- University of Bristol
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Education › Academic integrity and plagiarism
- Keywords
- generative AI, student perceptions, trust in AI outputs, AI-assisted learning, ChatGPT, plagiarism, reliability, gender differences, international students, artificial intelligence adoption, university students, inclusive pedagogy
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-06, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-06 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education. https://app.cochat.ai/papers/10.1080/13603108.2025.2595453
- The paper
- Cabrera, C., & Neville, R. (2025). Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education. Perspectives: Policy and Practice in Higher Education. https://doi.org/10.1080/13603108.2025.2595453
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.