The DOI matches its Crossref record; 53 of the 53 references it cites were found in a registry or an index; 22 of the 67 sources its claims cite were checked for support.
Addressing student non-compliance in AI use declarations: implications for academic integrity and assessment in higher education
Open accessAbstract
This study examines the factors driving student non-compliance with AI use declarations in academic assessments at King’s Business School, where 74% of students failed to declare AI usage despite declaration being a requirement of a mandatory coursework coversheet. Utilising the Theory of Planned Behaviour (TPB) as a framework, the research combines service evaluation survey data and semi-structured interviews to explore how attitudes, subjective norms, and perceived behavioural control influence compliance. Findings reveal that fear of academic repercussions, ambiguous guidelines, inconsistent enforcement, and peer influence are key barriers to AI use declaration. These factors complicate the declaration process, undermine transparency, and challenge academic integrity. The study extends the TPB model by highlighting the ethical and practical dilemmas posed by generative AI, which blur traditional norms of academic integrity. This research offers critical insights for policymakers, suggesting that clear, consistent, and trust-based policies are crucial in fostering ethical AI use. The findings underscore the importance of transparent communication and supportive institutional cultures to improve compliance. Ultimately, this study informs policy development by evaluating the effectiveness of declaration mechanisms and providing actionable recommendations to promote a culture of academic integrity in the evolving landscape of AI technologies.
Overview
CoChat Summary
TL;DR
This single-case mixed-methods study examines why students at King’s Business School failed to comply with a mandatory AI use declaration on coursework coversheets, despite up to 74% non-completion. Using the Theory of Planned Behaviour, it finds fear of academic consequences, ambiguous guidelines, inconsistent enforcement, and peer influence as key barriers to transparent AI declaration.
Key findings
- Up to 74% of students across undergraduate and postgraduate modules failed to complete the AI declaration, even though other coversheet sections were routinely filled out.
- In the survey, 57 students provided valid responses (3.7% of the population); 79% reported using AI tools such as ChatGPT, Grammarly, and Quillbot, but only 65% of those reported consistently completing the AI declaration.
- Thematic analysis of 17 interviews identified five themes: fear of academic consequences, ambiguity and lack of clarity in guidelines, belief that AI use should remain private, perceived inconsistency in enforcement, and peer influence/competition; “Everyone is using it” was echoed in 13 of the 17 interviews.
- The study extends the Theory of Planned Behaviour by showing that attitudes, subjective norms, and perceived behavioural control shape non-compliance, with AI declarations often viewed as risky self-incrimination rather than a neutral administrative task.
Methods
- Single case study at King’s Business School during the 2023–2024 academic year, using a sequential mixed-methods design: an anonymous service evaluation survey followed by semi-structured interviews.
- The survey was sent to all full-time undergraduate and postgraduate students via Moodle; it yielded 57 valid responses (3.7% of the population), with a £50 e-shopping voucher raffle incentive. Seventeen semi-structured interviews were conducted via Microsoft Teams, lasting 56 min to 1 h 43 min, with participants across programs and levels; participants received £30.
- Data were analysed using thematic analysis, with initial coding followed by theory-driven and emergent coding aligned with TPB dimensions; the sole researcher used multiple data readings, reflective journaling, and reflexivity.
Limitations
- Stated: the single-case design limits generalisability because of King’s Business School’s unique cultural and policy environment; findings are small-scale and context-specific, and the study captures only the early phase of AI policy implementation, so larger-scale, comparative, and longitudinal validation is needed.
- Evident from the text: the survey response rate was very low (57 valid responses, 3.7% of the population), and the sensitive topic may have biased self-reported AI use and declaration behaviour.
- Evident from the text: analysis was conducted by a sole researcher, which the paper addresses through reflexivity but still leaves potential for single-coder bias.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
18 Valid · 2 Not valid · 26 Not checked · 91 Own results and interpretations
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“TPB has been widely applied to academic integrity behaviours such as cheating and plagiarism (e.g. Passow et al. 2006; Stone, Jawahar, and Kisamore 2009; Imran and Nordin 2013; Khathayut, Walker-Gleaves, and Humble 2022).”
What the sources say
- Not checked — no full text of the sourcecites Passow et al. 2006 source
- Disputedcites Stone, Jawahar, and Kisamore 2009 sourceExperimental — accuracy not yet measured
In “The Theory of Planned Behavior and Academic Misconduct,” the paper lists studies applying TPB to cheating and plagiarism, but explicitly says its use in academic misconduct has been “fairly limited.” “Expanding the Theory of Planned Behavior for Academic Misconduct” likewise calls its use “very limited to date,” contrary to the claim that it has been widely applied in this area. Page numbers are not provided in our copy. (LLM, conf=0.88)
Checked by a reasoning model — may be wrong.
- Not checked — no full text of the sourcecites Imran and Nordin 2013 source
plain- 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.
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“This challenges TPB's assumption that individuals easily assess and control actions (Ajzen 1991, 2002), suggesting the model may be too simplistic for AI usage contexts.”
What the sources say
- Disputedcites Ajzen 1991 sourceExperimental — accuracy not yet measured
In “Predicting Behavior: Intentions and Perceived Behavioral Control” (page unavailable), Ajzen says most behaviors depend partly on opportunities and resources, not simply on an individual's ability to control them. The same section warns that perceived control may be a poor predictor when information is limited or a situation contains unfamiliar elements. This contradicts the assumption attributed to TPB; the application to AI usage is the citing authors’ inference. (LLM, conf=0.88)
Checked by a reasoning model — may be wrong.
- Not checked — no full text of the sourcecites Ajzen 2002 source
hedged- 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.
- Disputed
Every source the paper cites for this claim was checked and supports it.
In the paper
“Attitudes, defined as an individual's favourable or unfavourable evaluation of a specific behaviour (Ajzen 1991), are critical in determining whether students declare their use of AI tools in assessments.”
What the sources say
- Supportedcites Ajzen 1991 source
The citation is attached to the definition of attitudes, not the subsequent claim about students declaring AI-tool use. In “Affect versus Evaluation” (page unavailable), Ajzen describes an attitude toward a behavior as a reaction ranging from negative to positive favorability, supporting the cited definition. (LLM, conf=0.92)
plainstates a cause- 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.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Furthermore, the blurring of authorship resulting from human-AI collaborative writing complicates regulatory and pedagogical approaches in higher education (Kumar and Mindzak 2024).”
What the sources say
- Supportedcites Kumar and Mindzak 2024 source
The Discussion says GenAI blurs the line between individual work and assistance. The Literature Review says its availability raises questions about what universities consider misconduct, while the Conclusion calls for postsecondary institutions to re-examine assessments. Together, these support the paraphrase about regulatory and pedagogical challenges. (LLM, conf=0.90)
plainstates a cause- 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 cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This study also finds that non-compliance is often a deliberate decision driven by fear, with students viewing AI declarations as potential self-incrimination as opposed to a neutral administrative task (Stone 2023).”
What the sources say
- Partly supportedcites Stone 2023 sourceExperimental — accuracy not yet measured
Theme 1: Emotion reports fear of academic-integrity allegations, and Theme 2: Intentionality reports anxiety about being seen as a cheat. The paper does not report that students deliberately avoid AI declarations because they view them as self-incrimination; its “AI” refers to academic integrity, not artificial intelligence. (LLM, conf=0.86)
plainstates a cause- 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.
- Partly supported
63 more claims · 7 with their cited sources checked
Create a free account to see all 137 claims and every source verdict.
- Probably not a substantial claim. 39
- It may state two separate claims as one. 41
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Conversely, micro-level studies frequently focus on psychological determinants—such as technology readiness or anxiety—without accounting for the dynamic influence of macro-institutional environments and meso-level pedagogical interventions (Bai and Wang, 2025; Gonsalves, 2025).”
From Multidimensional responses to generative AI in higher education: an integrative framework linking institutional governance, teaching practice, and student psychological adaptation (2026) · cites this paper as Gonsalves, 2025
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Likewise, evidence suggests that the perception of punishment or stigma may encourage concealment strategies, fostering a low-trust environment and reinforcing control strategies rather than promoting learning-oriented ones [9,10].”
From Transparency Mechanisms for Generative AI Use in Higher Education Assessment: A Systematic Scoping Review (2022–2026) (2026) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“In such contexts, the absence of rubrics, checklists, or clear procedures can increase arbitrariness and evaluative bias, reinforcing concerns about fairness already highlighted by many authors [5,9].”
From Transparency Mechanisms for Generative AI Use in Higher Education Assessment: A Systematic Scoping Review (2022–2026) (2026) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“A central axis of variation is compliance assessment: some studies describe unverified self-disclosure, others indicate implicit instructor judgment in grading, and a minority propose explicit instruments (rubrics/checklists) or verifiable evidence of process [6,8,9,17].”
From Transparency Mechanisms for Generative AI Use in Higher Education Assessment: A Systematic Scoping Review (2022–2026) (2026) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This variability is closely tied to feasibility-perceived workload, acceptability, incentives for concealment, and surveillance risks [7,9-11].”
From Transparency Mechanisms for Generative AI Use in Higher Education Assessment: A Systematic Scoping Review (2022–2026) (2026) · cites this paper as [9]
Experimental — accuracy not yet measured
7 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Using thematic analysis in psychology
Braun, V., & Clarke, V. (2006) · Qualitative Research in Psychology · doi:10.1191/1478088706qp063oa · Cited by 194,414
The theory of planned behavior
Ajzen, I. (1991) · Organizational Behavior and Human Decision Processes · doi:10.1016/0749-5978(91)90020-t · Cited by 90,587
Research Design: Qualitative, Quantitative, and Mixed Methods Approaches
Creswell, J. W. (1994) · Cited by 35,473
Research design: qualitative, quantitative, and mixed methods approaches
Caldas, M. P. (2003) · Revista de Administração Contemporânea · doi:10.1590/s1415-65552003000100015 · Cited by 18,745 · cited 2×
“A sequential mixed-methods approach was used, incorporating an initial anonymous service evaluation survey (see Supplementary Material Appendix 2), followed by semi-structured inter-views (Creswell 2003).”
Perceived Behavioral Control, Self‐Efficacy, Locus of Control, and the Theory of Planned Behavior 1
Ajzen, I. (2002) · Journal of Applied Social Psychology · doi:10.1111/j.1559-1816.2002.tb00236.x · Cited by 7,629 · cited 2×
“When students feel capable of understanding and following the rules, they are more likely to comply (Ajzen 2002).”
Braun, V., & Clarke, V. (2023) · International Journal of Transgender Health · doi:10.1080/26895269.2022.2129597 · Cited by 2,878 · cited 1×
“Following explanation of the study aims, participants consented to video recording and anonymisation upon transcription. data were analysed using thematic analysis (Braun and Clarke 2006, 2023).”
Besanceney, P. H., Guetzkow, H., & White, W. R. (1952) · The American Catholic Sociological Review · doi:10.2307/3707197 · Cited by 1,440
Predicting dishonest actions using the theory of planned behavior
Beck, L., & Ajzen, I. (1991) · Journal of Research in Personality · doi:10.1016/0092-6566(91)90021-h · Cited by 1,325 · cited 3×
“This finding aligns with Beck and Ajzen (1991), which identified attitudes as a key predictor of dishonest actions, including cheating and lying to avoid academic responsibilities.”
Academic Dishonesty: Honor Codes and other Contextual Influences
McCabe, D. L., & Trevino, L. K. (1993) · The Journal of Higher Education · doi:10.1080/00221546.1993.11778446 · Cited by 855
ChatGPT in higher education: considerations for academic integrity and student learning
(2023) · Journal of Applied Learning & Teaching · doi:10.37074/jalt.2023.6.1.17 · Cited by 853 · cited 1×
“In response, universities have adopted varied approaches: some have banned generative AI or employed detection tools, while others permit and even encourage its use with appropriate acknowledgment (Perkins 2023; Sullivan, Kelly, and McLaughlan 2023; Xiao, Chen, and Bao 2023).”
Research works printed but not linked: 4
- Mulcahy S. and C. Goodacre. 2004. “Opening Pandora’s Box of Academic Integrity: Using Plagiarism Detection Software.” Paper presented at the Proceedings from ASCILITE Conference 2004.
- Ruona, W. E. A. 2005. “Analyzing Qualitative Data.” In Research in Organizations: Foundations and Methods of Inquiry, edited by Richard A. Swanson and Elwood F. Holton III, 233–263. San Francisco, CA: Berrett-Koehler Publishers.
- Tan C. M. M. Alhammad and M. Stelmaszak. 2024. “Teachers’ Perceptions of Students’ Use of Generative AI in Summative Assessments at Higher Education Institutions: An Exploratory Study.” Paper presented at the UK Academy for Information Systems Conference Proceedings 2024. https://aisel.aisnet.org/ukais2024/19?utm_source=aisel.aisnet.org/ukais2024/19&utm_medium=PDF&utm_campaign=PDFCoverPages.
- Yin, R. K. 2003. Case Study Research Design and Methods Third Edition, Applied Social Research Methods Series. California: Sage.
Cited by
Talk is cheap: why structural assessment changes are needed for a time of GenAI
Corbin, T., Dawson, P., & Liu, D. (2025) · Assessment & Evaluation in Higher Education · doi:10.1080/02602938.2025.2503964 · Cited by 107
“…that up to 74% of students did not complete the AI declaration appropriately, with many fearing that admitting AI use might be perceived negatively (Gonsalves 2024). many universities, including Cambridge and Princeton take a stricter approach, classifying unde-clared AI use as a form of…”
Oc, Y., Gonsalves, C., & Quamina, L. T. (2024) · Journal of Marketing Education · doi:10.1177/02734753241302459 · Cited by 112
Self-citationGenerative AI tools and empowerment in L2 academic writing
Moorhouse, B. L., Wan, Y., Wu, C., Wu, M., & Ho, T. Y. (2025) · System · doi:10.1016/j.system.2025.103779 · Cited by 63
Kangwa, D., Msafiri, M. M., & Fute, A. (2025) · Journal of Computer Assisted Learning · doi:10.1111/jcal.70109 · Cited by 22
Jayasinghe, S., Gamage, K. A. A., Yang, D., Cheng, C., Disanayake, C., & Apeji, U. D. (2026) · Education Sciences · doi:10.3390/educsci16010137 · Cited by 23
Metadata
- Authors
- Gonsalves, C.
- Affiliations
- King's College London (GB)
- Corresponding author
- Gonsalves, Chahna
- ORCID
- Gonsalves, Chahna
- Venue
- Assessment & Evaluation in Higher Education
- Publisher
- Informa UK Limited
- Published
- 2024-10-22
- Volume
- 50
- Issue
- 4
- Pages
- 592-606
- ISSN
- 0260-2938 (ISSN-L)
- 1469-297X
- OpenAlex
- W4403637141
- Cited by
- 212
- References
- 51
- Citations by year
- 2026: 171 · 2025: 39 · 2024: 2
- FWCI
- 84.98
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Assessment & Evaluation in Higher Education (publishedVersion, cc-by-nc-nd)
- Research Portal (King's College London) (submittedVersion, cc-by-nc-nd)
- Topics
- Social Sciences › Education › Academic integrity and plagiarism
- Social Sciences › Education › Artificial Intelligence in Education
- Social Sciences › Business, Management and Accounting › Organizational Behavior and Human Resource Management › Ethics in Business and Education
- Keywords
- generative AI, academic integrity, assessment, theory of planned behavior, peer influence, AI policy, higher education
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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (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: Addressing student non-compliance in AI use declarations: implications for academic integrity and assessment in higher education. https://app.cochat.ai/papers/10.1080/02602938.2024.2415654
- The paper
- Gonsalves, C. (2025). Addressing student non-compliance in AI use declarations: implications for academic integrity and assessment in higher education. Assessment & Evaluation in Higher Education. https://doi.org/10.1080/02602938.2024.2415654
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.