From avoidance to adoption: strategies for integrating AI in higher education contexts
Open accessThe DOI matches its Crossref record; 29 of the 30 references it cites were found in a registry or an index; 9 of the 45 sources its claims cite were checked for support.
Abstract
Introduction: This paper aims to identify effective strategies for integrating Artificial Intelligence (AI) in higher education (HE), specifically addressing the challenges related to AI literacy, ethical concerns, and institutional variations. It provides a framework for incorporating an “embracing” approach to AI adoption, characterised by strategic incorporation of AI tools and underpinned by AI literacy and constructive AI skepticism. Methods: The paper employs three main approaches. First, we explore current challenges faced by instructors in taking an embracing approach, particularly in relation to AI literacy. Second, we introduce instructor AI skepticism from a multi-layered perspective. Third, we present a series of case studies from universities in Northern Ireland reflecting on challenges of the embracing approach across STEM, Humanities, and Social Sciences, and how these challenges have been addressed. Results: An exploration of AI adoption across different subject areas in four case studies demonstrates substantial variability in AI literacy, AI skepticism and acceptance across disciplines. Nonetheless, broader common themes emerge across the case studies. These themes generally involve critical thinking and application of subject area-specific theories to new contexts. Crucially, structured instructional approaches significantly reduce misconceptions and enhance ethical AI usage. Discussion: This study uniquely integrates discipline-specific case studies and institutional context analyses. It contributes to understanding AI literacy as a multidimensional skillset. It further positions AI literacy as a critical means of addressing the underrepresented instructors' multilayered AI skepticism, with an emphasis on ethical considerations and critical student engagement. Our recommendations highlight the necessity for targeted faculty training, clear institutional policies, and structured curriculum integration strategies to foster AI literacy, address instructor skepticism, and enhance ethical awareness effectively.
Overview
CoChat Summary
TL;DR
This conceptual and case-study paper argues that higher education should move from an "avoidance" stance toward an "embracing" approach to generative AI, positioning AI literacy as the means of addressing instructors' multi-layered AI skepticism. Four discipline-specific case studies from Northern Ireland universities (Engineering, Linguistics, NLP, and medieval Literature) illustrate how structured comparison of student work against AI output can build critical evaluation, ethical reasoning, and creative application.
Key findings
- Across four case studies, AI literacy, AI skepticism, and acceptance varied substantially by discipline, but common themes emerged: critical thinking and application of subject-specific theories to new contexts; structured instructional approaches "significantly reduce misconceptions and enhance ethical AI usage."
- Instructor AI skepticism is layered across three levels: cross-profession (ethical concerns, data privacy, bias), educational (pedagogical concerns about quantification and transmissive instruction), and disciplinary (STEM instructors doubt AI's conceptual depth; Humanities instructors doubt its interpretive nuance).
- The paper reframes skepticism as constructive: "constructive instructor AI skepticism functions as the motivational driver for designing AI literacy-oriented pedagogies, whereas instructor AI literacy provides the pedagogical, disciplinary, and ethical competencies required to implement these pedagogies."
- AI literacy is proposed as both a shared cross-disciplinary foundation and a discipline-specific construct; the case studies show Engineering prioritised technical reliability and verification, Linguistics evaluated AI explanations against theory, NLP centred on model interpretability, and Literature on interpretive complexity and authorship.
Methods
- Three approaches: (1) exploration of instructor challenges in taking an embracing approach, particularly AI literacy; (2) introduction of a multi-layered framework of instructor AI skepticism (cross-profession, educational, disciplinary); (3) four case studies from undergraduate and postgraduate modules at Ulster University and Queen's University Belfast, spanning STEM (Engineering, Computing/NLP), Social Sciences (Linguistics), and Humanities (English).
- Case studies describe classroom activities: Engineering (Mechatronics 1, Engineering Programming) used step-by-step tutorials, MATLAB AI apps, AI-citation guidelines, and hybrid AI-plus-traditional analysis; Linguistics (Language Acquisition) and NLP (Deep Learning & NLP) used a three-step sequence — define/apply concept without AI, generate AI output, compare; Literature (Chaucer's *Canterbury Tales*) used Voyant, AntConc, and ChatGPT outputs as seminar discussion prompts.
- Analysis is qualitative and reflective; the paper reports no quantitative measures, and the case studies were selected from materials used on the authors' own taught modules.
Limitations
- The paper states that the Engineering case study avoided "the assumption that confidence, AI literacy, or ethical awareness had been formally measured," and that future delivery could strengthen the approach with a clearer AI adaptation framework.
- The Literature case study notes that "how students experienced these approaches remains an open question, and one worth pursuing."
- Evident from the text (not stated as a limitation by the authors): the case studies are drawn from the authors' own modules at two Northern Irish institutions, with no formal student data, comparison group, or outcome measurement, so the claimed reductions in misconceptions and gains in ethical awareness rest on instructor observation rather than systematic evidence.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
4 Valid · 7 Not valid · 26 Not checked · 106 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“At the broadest level, skepticism toward AI is not confined to education but is shared across professional domains such as medicine, law, and economics.”
What the sources say
- Partly supportedcites Adepoju et al., 2024 sourceExperimental — accuracy not yet measured
Section 3 (page unavailable) discusses skepticism toward technology in workplaces, supporting the point that such skepticism is not confined to education. The Introduction discusses concerns about AI in labor economics, and Section 8 discusses AI’s implications for healthcare and law, but the copy does not establish that skepticism toward AI is shared across all three named domains. (LLM, conf=0.82)
- Partly supportedcites Ambartsoumean and Yampolskiy, 2023 sourceExperimental — accuracy not yet measured
The Introduction discusses skepticism about AI risks, and the Gatekeeping section describes skeptical arguments involving computer scientists and engineers. That supports the broader point that the discussion is not confined to education, but the cited passages do not establish that skepticism toward AI is shared across medicine, law, and economics. The paper also distinguishes skepticism about AI risks from skepticism toward AI generally. (LLM, conf=0.79)
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.
- Partly supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Kong et al., (2025) show that in STEM domains, AI feedback is often accurate but lacks conceptual depth or flexibility.”
What the sources say
- Could not verifycites Kong et al., 2025 sourceExperimental — accuracy not yet measured
The section “Possibilities of AI education and AI literacy” (page not available) says AI can provide timely feedback, but does not assess its accuracy or conceptual depth. A search for STEM feedback and its accuracy, conceptual depth, and flexibility found no passage addressing those claims in our copy. (LLM, conf=0.72)
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.
- Could not verify
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“In these contexts and others, skepticism often centers on ethical concerns that take many shapes and forms (Adepoju et al., 2024; Ambartsoumean and Yampolskiy, 2023).”
What the sources say
- Partly supportedcites Adepoju et al., 2024 sourceExperimental — accuracy not yet measured
Section 3, “Human Skepticism on Modern Technologies in the Workplace” (page unavailable), links workplace skepticism to concerns about job security, work-life balance, and personal privacy. Privacy is an ethical concern, but the paper does not establish that skepticism often centers on ethical concerns or characterize them as taking many forms. (LLM, conf=0.78)
- Partly supportedcites Ambartsoumean and Yampolskiy, 2023 sourceExperimental — accuracy not yet measured
Our copy does treat ethical concerns as a recurring cluster of AI-risk skepticism: the 'Something Else Takes Priority' section lists skeptics pointing to 'unethical robots, mistakes with self-driving cars, algorithmic and data bias, privacy concerns, technological unemployment... limited transparency, surveillance' and current military applications, and 'We Are Unsure About Something, Thus Don't Worry' notes that 'ethical concerns, morality, and consciousness' are a concrete part of AI Safety research. Professional/business settings appear too (businesses claiming to be 'responsible actors' so as not to need regulation; AI review boards weighing 'harms, societal impacts, and ethical assumptions'). What the source does not support is the claim's centrality and scope qualifiers: the survey explicitly categorizes skepticism 'by the type of mistaken thinking involved' (gatekeeping/ad hominem/authority bias, 'something is impossible illusion', giving up on even trying, large assumptions of faith, negative anthropocentrism, using history to predict the future), so ethical concerns are one species among many rather than what skepticism 'often centers on', and the paper offers no account of skepticism in 'professional contexts' as such. (LLM, conf=0.68)
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.
- Partly 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
“the case studies show how 'AI literacy-oriented' learning activities can support the development of students' AI literacy by fostering critical evaluation, ethical reasoning, and creative application”
plainstates a causen = 4- 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.
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
“Such tools enabled students to quickly identify recurring lexical patterns and quickly identify frequent occurrences of terms such as 'gentillesse' and 'pilgrimage'.”
plainstates a causeSingle case- 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.
55 more claims
Create a free account to see all 143 claims and to verify the 1 reference still open.
- It may state two separate claims as one. 47
- Probably not a substantial claim. 51
- The quote only partly states this claim. 2
Citations
Cites
Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-019-0171-0 · Cited by 7,242
What is AI Literacy? Competencies and Design Considerations
Long, D., & Magerko, B. (2020) · Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376727 · Cited by 3,241
The Structural Sources of Verb Meanings
Gleitman, L. (1990) · Language Acquisition · doi:10.1207/s15327817la0101_2 · Cited by 1,783
The social power of algorithms
Beer, D. (2017) · Information, Communication & Society · doi:10.1080/1369118x.2016.1216147 · Cited by 945
The Cambridge Handbook of the Learning Sciences
(2014) · doi:10.1017/cbo9781139519526 · Cited by 907
More than tools? Making sense of the ongoing digitizations of higher education
Castañeda, L., & Selwyn, N. (2018) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-018-0109-y · Cited by 445
Understanding Pedagogy and its Impact on Learning
Mortimore, P. (1999) · SAGE Publications Ltd eBooks · doi:10.4135/9781446219454 · Cited by 324
Digital Scholarly Editing: Theories and Practices
Driscoll, M. J., Pierazzo, E., & body, O. B. P. I. (2016) · Open Book Publishers · doi:10.11647/obp.0095 · Cited by 234
Unveiling the shadows: beyond the hype of AI in education
Al-Zahrani, A. M. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e30696 · Cited by 225
Automation: is it really different this time?
Wajcman, J. (2017) · The British Journal of Sociology · doi:10.1111/1468-4446.12239 · Cited by 220
Towards social generative AI for education: theory, practices and ethics
Sharples, M. (2023) · Learning: Research and Practice · doi:10.1080/23735082.2023.2261131 · Cited by 217
Automation, APIs and the distributed labour of platform pedagogies in google classroom
Perrotta, C., Gulson, K. N., Williamson, B., & Witzenberger, K. (2021) · Critical Studies in Education · doi:10.1080/17508487.2020.1855597 · Cited by 208
ENAI Recommendations on the ethical use of Artificial Intelligence in Education
Foltynek, T., Bjelobaba, S., Glendinning, I., Khan, Z. R., Santos, R., Pavletic, P., & Kravjar, J. (2023) · International Journal for Educational Integrity · doi:10.1007/s40979-023-00133-4 · Cited by 180
Ethics of AI: a systematic literature review of principles and challenges
Khan, A. A., Badshah, S., Liang, P., Waseem, M., Khan, B., Ahmad, A., Fahmideh, M., Niazi, M., & Akbar, M. A. (2022) · The International Conference on Evaluation and Assessment in Software Engineering 2022 · doi:10.1145/3530019.3531329 · Cited by 170
The Generative AI Paradox: "What It Can Create, It May Not Understand"
West, P., Lu, X., Dziri, N., Brahman, F., Li, L., Hwang, J. D., Jiang, L., Fisher, J., Ravichander, A., Chandu, K. R., et al. (2023) · International Conference on Learning Representations · doi:10.48550/arxiv.2311.00059 · Cited by 122
Artificial Intelligence (AI) literacy—An argument for AI literacy in education
Kong, S.-C., Korte, S.-M., Burton, S., Keskitalo, P., Turunen, T., Smith, D., Wang, L., Lee, J. C.-K., & Beaton, M. C. (2025) · Innovations in Education and Teaching International · doi:10.1080/14703297.2024.2332744 · Cited by 65
Williamson, B. (2019) · Rethinking Pedagogy for a Digital Age · doi:10.4324/9781351252805-14 · Cited by 62
Daskalaki, E., Psaroudaki, K., & Fragopoulou, P. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2408.15686 · Cited by 58
AI In education: comparative perspectives from STEM and non-STEM instructors
Parviz, M. (2024) · Computers and Education Open · doi:10.1016/j.caeo.2024.100190 · Cited by 50
Modelling in digital humanities: signs in context
Ciula, A., & Eide, Ø. (2016) · Digital Scholarship in the Humanities · doi:10.1093/llc/fqw045 · Cited by 40
Virtual Laboratories in STEM Higher Education: A Scoping Review
Sellberg, C., Nazari, Z., & Solberg, M. (2024) · Nordic Journal of Systematic Reviews in Education · doi:10.23865/njsre.v2.5766 · Cited by 39
The nature, causes, and effects of skepticism on technology diffusion
Trieste, L., & Turchetti, G. (2024) · Technological Forecasting and Social Change · doi:10.1016/j.techfore.2024.123663 · Cited by 20
International Journal for Digital Society
(2016) · International Journal for Digital Society · doi:10.20533/ijds.2040.2570 · Cited by 17
Critical Digital Humanities: The Search for a Methodology
DOBSON, J. E. (2019) · doi:10.5406/j.ctvfjd0mf · Cited by 13
AI Risk Skepticism, A Comprehensive Survey
Ambartsoumean, V. M., & Yampolskiy, R. V. (2023) · arXiv · doi:10.48550/arxiv.2303.03885 · Cited by 10
Artificial intelligence skepticism in career domains
Adepoju, O. D., Tijani, B., & Karera, S. (2024) · International Journal for Digital Society · doi:10.20533/ijds.2040.2570.2024.0236 · Cited by 6
Ziemińska, R. (2017) · doi:10.3726/978-3-653-04808-7 · Cited by 3
L’atelier de Christine de Pizan by Inès Villela-Petit
Walters, L. J. (2021) · Digital philology · doi:10.1353/dph.2021.0019 · Cited by 1
Barbero (2014) · doi:10.1016/j · Cited by 1
Echeverri, S. (2025) · doi:10.21428/e2759450.95725785 · Cited by 1
Research works printed but not linked: 7
- 2022, Economic and Digital Infrastructure Strategy
- 2021, AI in Healthcare: Framework for Ethical Implementation
- Echeverri, 2025, Skepticism. in Open Encyclopedia of Cognitive Science
- Luckin, 2016, Intelligence Unleashed: An Argument for AI in Education
- Selwyn, 2019, Should Robots Replace Teachers? AI and the Future of Education
- 2024, Guidance for Students on the Use of AI Tools
- 2024, AI competency Framework for Teachers
Metadata
- Authors
- Gerard, J., Hadi, M. U., Natsheh, B. R. R., Singh, M., & Reid, A.
- Affiliations
- University of Ulster (GB)
- School of Arts, English and Languages, Queen's University Belfast
- Corresponding author
- Gerard, Juliana
- Venue
- Frontiers in Education
- Publisher
- Frontiers Media SA
- Published
- 2026-08-12
- Volume
- 11
- ISSN
- 2504-284X (ISSN-L)
- OpenAlex
- W7202267311
- Semantic Scholar
- 659f188def915dfab4bece70892440258ebfc0f8
- Cited by
- 0
- References
- 21
- FWCI
- 0.00
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Frontiers in Education (publishedVersion, cc-by)
- Ulster University Research Portal (Ulster University) (acceptedVersion, other-oa)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Education › Digital literacy in education
- Keywords
- artificial intelligence in higher education, AI literacy, artificial intelligence adoption, AI ethics, faculty training, curriculum integration, disciplinary differences, institutional AI policies, critical thinking
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 and 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
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.