Student or educator? Sign up with your school email and get 6 months of Pro free.
Cited by 403Artificial Intelligence in EducationArtificial Intelligence in Healthcare and EducationAcademic integrity and plagiarism

The DOI matches its Crossref record; 32 of the 32 references it cites were found in a registry or an index; 4 of the 79 sources its claims cite were checked for support.

Investigating the higher education institutions’ guidelines and policies regarding the use of generative AI in teaching, learning, research, and administration

Open access
An, Yunjo, Yu, Ji Hyun, James, Shadarra·2025·International Journal of Educational Technology in Higher Education

Abstract

This study examined the guidelines issued by the top 50 U.S. universities regarding the use of Generative AI (GenAI) in academic and administrative activities. Employing a mixed methods approach, the research combined topic modeling, sentiment analysis, and qualitative thematic analysis to provide a comprehensive understanding of institutional responses to GenAI. Topic modeling identified four core topics: Integration of GenAI in Learning and Assessment, GenAI in Visual and Multimodal Media, Security and Ethical Considerations in GenAI, and GenAI in Academic Integrity. These themes were further explored through sentiment analysis, which revealed highly positive attitudes towards GenAI across all institution types, with significant differences between faculty and student-targeted guidelines. Qualitative thematic analysis corroborated these findings and provided deeper insights, revealing that 94% of universities had faculty guidelines emphasizing the importance of establishing and communicating course-specific GenAI policies. This analysis also highlighted recurring themes such as academic integrity and privacy concerns, which aligned with the security and ethical considerations identified in the topic modeling. The study highlights the rapid evolution of GenAI guidelines in higher education and the need for flexible, stakeholder-specific policies that address both the opportunities and challenges presented by this technology.

Overview

CoChat Summary

TL;DR

This study analyzed publicly available generative AI guidelines from the top 50 U.S. universities using topic modeling, sentiment analysis, and qualitative thematic analysis of 214 documents. It found four dominant topics and highly positive sentiment overall, with significant differences by target audience and uneven coverage across stakeholder groups.

Key findings

  • LDA identified four topics: Integration of GenAI in Learning and Assessment (30% of tokens), GenAI in Visual, Interactive, and Multimodal Media (28.8%), Security and Ethical Considerations in GenAI (21.3%), and GenAI in Academic Integrity (20%).
  • Sentiment was highly positive across institution types: private institutions mean = 0.945355, public = 0.963464, with no significant difference (Mann-Whitney U = 5664.5000, p = 0.9664). Audience differences were significant (Kruskal-Wallis = 22.7987, p = 0.0004), with a significant faculty-vs-student difference (Dunn’s p = 0.000460); administrators had the highest mean (0.996750), while researchers (0.898900) and students (0.910272) were lower.
  • Faculty guidelines were most common: 47 of 50 universities (94%) provided them; 56% offered syllabus statement guidance, 54% advised setting and communicating course policy, 48% advised redesigning assignments/assessments, and 38% discouraged AI detection tools.
  • Coverage varied by audience: 34 of 50 universities (70%) had all-stakeholder guidelines, 21 (42%) had student guidelines, 9 (18%) had researcher guidelines, and 7 (14%) had staff/administrator guidelines.

Methods

  • Data: top 50 U.S. universities from Times Higher Education’s 2024 list; manual searches of official websites using GenAI/ChatGPT guideline and policy keywords; inclusion criteria were university-wide, publicly available, English-language guidelines. The corpus contained 214 documents, 9,475 sentences, and 235,118 words.
  • Analysis: a four-step procedure of data collection, topic modeling, sentiment analysis, and qualitative thematic analysis. Topic modeling used TF/TF-IDF and optimized LDA; sentiment analysis used NLTK’s VADER Sentiment Intensity Analyzer compound scores from -1 to +1; qualitative thematic analysis coded guidelines by target audience, with two researchers coding a subset and reaching consensus.
  • Statistical comparisons included normality tests, Mann-Whitney U tests, Kruskal-Wallis tests, and post-hoc Dunn’s tests.

Limitations

  • Stated: the study reviewed only publicly available guidelines from the top 50 U.S. universities, so non-public guidelines were excluded and findings may not generalize to other U.S. or international institutions.
  • Stated: the number of guidelines for staff and administrators was relatively small, and data were collected during Spring 2024; because GenAI guidelines evolve, longitudinal and cross-cultural research is needed.
  • Evident: the analysis captures guideline text rather than implementation, compliance, or effects on pedagogy and student outcomes; VADER was designed for social media sentiment and adapted to academic texts; LDA topic number and labels were determined by coherence scores and domain knowledge, so interpretations may be sensitive to modeling choices.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1186/s41239-025-00507-3

Cites

32 of 32 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 Study procedure
Fig. 1 Study procedureCC-BY · From the PDF
Fig. 2 Coherence scores by number of topics
Fig. 2 Coherence scores by number of topicsCC-BY · From the PDF
Fig. 3 4 Topic Model
Fig. 3 4 Topic ModelCC-BY · From the PDF

Verification

Claims the paper makes

3 Valid · 1 Not valid · 54 Not checked · 96 Own results and interpretations

59 more claims

Create a free account to see all 154 claims and to verify the 61 references still open.

See all claims — free account
  • Probably not a substantial claim. 70
  • It may state two separate claims as one. 36

Create a free account to see these claims.

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.

2 of 2

Citations

Cites

Show all 28 references

Research works printed but not linked: 2

  1. Miles, M.B., Huberman, A.M., & Saldana, J. (2014). Qualitative data analysis: A methods sourcebook (3rd ed.) Sage.
  2. Sleator, L. (2023). Almost half of Cambridge students admit they have used ChatGPT. The Times. https://www.thetimes.com/business-money/technology/article/cambridge-university-students-chatgpt-ai-degree-2023-rnsv7mw7z

Cited by

See all 403

Metadata

Authors
An, Y., Yu, J. H., & James, S.
Affiliations
University of North Texas (US)
Corresponding author
An, Yunjo
Venue
International Journal of Educational Technology in Higher Education
Published
2025-02-21
Volume
22
Issue
1
ISSN
2365-9440 (ISSN-L)
OpenAlex
W4407778399
Cited by
403
References
28
Citations by year
2026: 319 · 2025: 84
FWCI
321.82
Type
journal article
Language
en
Open access
Licence
Locations
  • International Journal of Educational Technology in Higher Education (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
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, higher education institutions, institutional policy, AI guidelines, academic integrity, AI-assisted assessment, multimodal media, AI ethics, data security, privacy concerns, American universities

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 2026-11-13, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-13 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash).

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: Investigating the higher education institutions’ guidelines and policies regarding the use of generative AI in teaching, learning, research, and administration. https://app.cochat.ai/papers/10.1186/s41239-025-00507-3
The paper
An, Y., Yu, J. H., & James, S. (2025). Investigating the higher education institutions’ guidelines and policies regarding the use of generative AI in teaching, learning, research, and administration. International Journal of Educational Technology in Higher Education. https://doi.org/10.1186/s41239-025-00507-3

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.