The DOI matches its Crossref record; 66 of the 71 references it cites were found in a registry or an index; 23 of the 49 sources its claims cite were checked for support.
Developing the HAKI model as a conceptual framework for AI-assisted academic research
Open accessAbstract
The rapid adoption of artificial intelligence (AI) in academia has transformed knowledge production while raising critical questions regarding authorship, academic integrity, and ethical responsibility. Despite the growing integration of AI into research practices, there remains a lack of conceptual frameworks that clearly define the boundaries of ethical human–AI collaboration in academic research. Addressing this gap, the present study explores academics’ perceptions of AI through semi-structured interviews with 131 faculty members from Iranian and Canadian universities representing diverse academic disciplines. The interview data were analyzed using thematic analysis. The findings indicate that although AI enhances research efficiency and supports various stages of the research process, participants expressed significant concerns regarding algorithmic bias, academic integrity, transparency, and the potential erosion of scholarly voice and intellectual ownership. A dominant finding was the perceived need for a structured framework to guide the ethical integration of AI into academic research. In response, this study proposes the HAKI framework, a conceptual model that emphasizes human oversight, transparency, critical interpretation, and recursive verification to support responsible AI-assisted knowledge production. As a conceptual contribution, the HAKI framework is intended to provide a theoretical foundation for future empirical investigation rather than a validated operational model. The study further highlights the importance of institutional governance, ethical guidelines, and AI literacy initiatives to facilitate responsible AI adoption in higher education and scholarly research.
Overview
CoChat Summary
TL;DR
This qualitative interview study of 131 faculty in Iran and Canada finds AI is widely used for research efficiency but raises concerns about bias, unreliable citations, academic integrity, accountability, and homogenized scholarly voice. It proposes the HAKI conceptual framework—Human-Centric Knowledge, AI-Assisted Knowledge, and Ethical-Regulatory Oversight, coordinated through Human Analytical Control, AI Ethical Mediation, Critical Interpretation, and Recursive Verification—as a foundation for future empirical validation.
Key findings
- Participants largely viewed AI as an efficiency tool: 87.8% reported AI increases efficiency, 74.0% saw it as time-saving, and 67.2% said it aids literature review; yet 70.2% said AI lacks depth in analysis, 50.4% found AI-generated citations often unreliable, and 31.3% said AI may reinforce biases.
- Ethical concerns were prominent: 83.2% reported AI-generated plagiarism concerns, 74.8% cited lack of accountability for AI outputs, 65.6% called for clear AI disclosure policies, and 47.3% said AI reduces students’ academic integrity.
- On authorship and voice, 67.9% said AI homogenizes academic voice, 55.0% said AI-generated content lacks personal touch, and 38.9% feared AI replacing human authorship.
- Nearly all participants saw a need for a new framework: 96.9% endorsed the need for a new AI-human knowledge framework, 80.2% emphasized human-AI collaboration, 67.2% wanted AI transparency guidelines, and 69.4% said universities should provide AI training.
Methods
- Qualitative design using semi-structured interviews with 131 university faculty members (73 men, 58 women) from five universities and colleges in Iran and Canada, recruited purposively; inclusion required a current academic appointment and self-reported AI use in research writing within the previous 12 months.
- Interviews lasted approximately 20–30 minutes, conducted via Zoom or in person, primarily in English (n = 119) with a subset in Farsi (n = 12); Farsi transcripts were translated and back-checked.
- Data were analyzed using Braun and Clarke’s six-phase reflexive thematic analysis, combining deductive and inductive coding; two analysts independently double-coded 20% of transcripts with 85% agreement, member checking was conducted with a subset, and sampling continued until thematic saturation. The HAKI framework was developed conceptually, grounded in Sociotechnical Systems Theory and Epistemic Agency Theory.
Limitations
- The paper states that HAKI remains a conceptual and exploratory model rather than an empirically validated framework; it assumes human oversight can effectively regulate AI-generated knowledge, may be difficult to implement across diverse institutions, assumes an AI-literate scholarly community, and does not fully account for evolving generative AI capabilities.
- The paper states that purposive recruitment through professional networks and the researcher’s academic affiliations may have introduced selection bias, and findings should be interpreted as analytically transferable rather than statistically generalizable.
- Evident from the text: the sample is limited to faculty in Iran and Canada, relies on self-reported AI use, uses relatively short interviews (≈20–30 minutes), and does not empirically test or validate the proposed HAKI framework.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
15 Valid · 7 Not valid · 14 Not checked · 130 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Actor-Network Theory (ANT) [15, 50, 51] conceptualizes technologies as active participants within sociotechnical networks, but it attributes a level of agency to AI systems that cannot exercise consciousness, moral judgment, or accountability [11].”
What the sources say
- Could not verifycites [11] Bojic, S., Radovanovic, N., Radovic, M., & Stamenkovic, D. (2024). Could generative artificial intelligence replace fieldwork in pain research? Scandinavian Journal of Pain, 24(1). https://doi.org/10.1515/sjpain-2023-0136 sourceExperimental — accuracy not yet measured
The Introduction and Discussion examine whether generative AI can provide reliable pain-research data, but do not discuss Actor-Network Theory or agency. Searches for consciousness, moral judgment, accountability, and sociotechnical networks found no passage addressing the attributed claim. (LLM, conf=0.74)
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
“HAKI's approach aligns with emerging guidelines from leading academic publishers, which emphasize AI disclosure and authorship clarity [86].”
What the sources say
- Partly supportedcites [86] van Dis EAM, Bollen J, Zuidema W, van Rooij R, Bockting CL. ChatGPT: five priorities for research. Nature. 2023;614(7947):224–6. https://doi.org/10.1038/d41586-023-00288-7. sourceExperimental — accuracy not yet measured
In “Develop rules for accountability,” the paper recommends disclosing AI use in author-contribution statements and acknowledgements, and says LLMs should not be listed as authors. It calls for publishers to adopt explicit policies; it does not establish that leading academic publishers had already issued the guidelines described in the claim. (LLM, conf=0.86)
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.
- Partly supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“However, many researchers may lack sufficient AI literacy [16, 70] to fully understand how AI algorithms function, where biases originate, and how AIgenerated knowledge is structured.”
What the sources say
- Could not verifycites [16] Chaudhry S, Muhammad A, Shah AA, Batoo F. Human-computer user interface design for semiliterate and illiterate users. Lahore Garrison Univ Res J Comput Sci Inf Technol. 2021;5(4):62–77. sourceExperimental — accuracy not yet measured
The Introduction and Conclusions discuss ICT literacy and interface design for semiliterate and illiterate users, not researchers’ AI literacy. Searches for AI algorithms, bias, and AI-generated knowledge found no passage addressing the attributed claim; the copy is not truncated. (LLM, conf=0.76)
- Not checked — no full text of the sourcecites [70] Seker O, Kwon K, Kocak O. Exploring researchers’ artificial intelligence (AI) literacy: the mediating role of digital literacy and data literacy between 21st century skills and AI literacy. Inf Dev. 2025. https://doi.org/10.1177/02666669251336368. 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.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Although numerous studies have investigated faculty perceptions of AI in higher education and academic writing, most have focused on documenting benefits, risks, or ethical concerns [21, 92] and less attention has been devoted to translating these perspectives into an integrated conceptual framework that guides ethical human-AI collaboration throughout the research process.”
What the sources say
- Could not verifycites [21] Crompton H, Burke D. Artificial intelligence in higher education: the state of the field. Int J Educ Technol High Educ. 2023;20:22. https://doi.org/10.1186/s41239-023-00392-8. sourceExperimental — accuracy not yet measured
The Introduction discusses benefits of AI for instructors, and Subject domains mentions AI used for writing, but neither addresses studies of faculty perceptions. The review’s Gaps and future research section says little work examined how AI assists instructors; it does not establish that most faculty-perception studies focused on benefits, risks, or ethical concerns. Searched the indexed copy for faculty perceptions, attitudes, risks, and ethics. (LLM, conf=0.76)
- Could not verifycites [92] Zawacki-Richter O, Marín VI, Bond M, Gouverneur F. Systematic review of research on artificial intelligence applications in higher education – where are the educators? Int J Educ Technol High Educ. 2019;16:39. https://doi.org/10.1186/s41239-019-0171-0. sourceExperimental — accuracy not yet measured
The Introduction and Methods describe a review of AI applications in higher education, not a review of studies on faculty perceptions or academic writing. The section “Understanding of AI and critical reflection of challenges and risks” reports that only two of 146 articles critically addressed ethics and risks, but it does not establish what most faculty-perception studies focused on. Searches for faculty perceptions and academic writing found no passage addressing that claim. (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
“Although numerous studies have investigated faculty perceptions of AI in higher education and academic writing, most have focused on documenting benefits, risks, or ethical concerns [21, 92] and less attention has been devoted to translating these perspectives into an integrated conceptual framework that guides ethical human-AI collaboration throughout the research process.”
What the sources say
- Could not verifycites [21] Crompton H, Burke D. Artificial intelligence in higher education: the state of the field. Int J Educ Technol High Educ. 2023;20:22. https://doi.org/10.1186/s41239-023-00392-8. sourceExperimental — accuracy not yet measured
The Introduction identifies a gap in understanding AI’s collective affordances in higher education, and Gaps and future research lists areas needing further study. Neither section addresses whether conceptual frameworks define the boundaries of ethical human–AI collaboration throughout academic research; a search for that framework and for faculty perceptions and ethical concerns found no such passage. In the citing sentence, the framework point also follows the citations rather than being directly attached to them. (LLM, conf=0.74)
- Partly supportedcites [92] Zawacki-Richter O, Marín VI, Bond M, Gouverneur F. Systematic review of research on artificial intelligence applications in higher education – where are the educators? Int J Educ Technol High Educ. 2019;16:39. https://doi.org/10.1186/s41239-019-0171-0. sourceExperimental — accuracy not yet measured
In “Understanding of AI and critical reflection of challenges and risks,” the review finds that only two of 146 articles critically reflect on ethical implications of AI in education. Its discussion also identifies a lack of theory in AI-driven educational technology. These findings support a broad gap in ethical and theoretical work, but the paper does not establish a lack of frameworks defining ethical human–AI collaboration throughout the academic research process; no page numbers are available in our copy. (LLM, conf=0.80)
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
67 more claims · 5 with their cited sources checked
Create a free account to see all 166 claims and to verify the 4 references still open.
- It may state two separate claims as one. 46
- Probably not a substantial claim. 75
- The quote only partly states this claim. 3
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 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
§ 3 Theoretical framework · p. 4
“Human-AI Collaboration (HAC), AI Ethical Mediation (AEM), Cognitive Integration (CI), and Recursive Verification (RV)”
§ 6.2 Operational components of the HAKI framework · p. 19
“Human Analytical Control (HAC)”
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
§ 5.3 Theme 3: ethical concerns and academic integrity · p. 11
“Table 4 summarizes the principal ethical concerns identified by participants.”
§ "...but when it generates references, they are often made up." · p. 9
“Table 4 the most frequently reported advantages and limitations of AI identified by the participants”
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
§ 3 Theoretical framework · p. 4
“Human-AI Collaboration (HAC), AI Ethical Mediation (AEM), Cognitive Integration (CI), and Recursive Verification (RV)”
§ 6.2 Operational components of the HAKI framework · p. 19
“Critical Interpretation (CI)”
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
§ 4.1 Participants · p. 5
“all analytical decisions, coding, and thematic interpretations were conducted by the researcher”
§ 4.3 Data collection and analysis · p. 7
“Two analysts independently double-coded 20% of transcripts”
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 Discovery of Grounded Theory; Strategies for Qualitative Research
Glaser, B. G., Strauss, A. L., & Strutzel, E. (1968) · Nursing Research · doi:10.1097/00006199-196807000-00014 · Cited by 39,449
How Many Interviews Are Enough?: An Experiment with Data Saturation and Variability
Guest, G., Bunce, A., & Johnson, L. (2006) · Field Methods · doi:10.1177/1525822x05279903 · Cited by 20,808
The Knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation
Nonaka, I., & Takeuchi, H. (1995) · doi:10.1093/oso/9780195092691.001.0001 · Cited by 18,010
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 · cited 1×
“Their day-to-day engagement with AI-assisted research places them in a unique position to evaluate not only the practical affordances of AI but also its broader epistemological and ethical implications for knowledge production [10, 26, 44].”
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 · cited 1×
“…faculty perceptions of AI in higher education and academic writing, most have focused on documenting benefits, risks, or ethical concerns [21, 92] and less attention has been devoted to translating these perspectives into an integrated conceptual framework that guides ethical human–AI…”
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988
Clark, A., & Chalmers, D. (1998) · Analysis · doi:10.1093/analys/58.1.7 · Cited by 6,333 · cited 1×
“Extended Mind Theory (EMT) [20] views technology as an extension of human cognition, however, it does not adequately distinguish between passive cognitive tools and autonomous generative AI systems capable of producing original textual outputs [64].”
Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., Baabdullah, A. M., Koohang, A., Raghavan, V., Ahuja, M., et al. (2023) · International Journal of Information Management · doi:10.1016/j.ijinfomgt.2023.102642 · Cited by 4,584
Notes on the theory of the actor-network: Ordering, strategy, and heterogeneity
Law, J. (1992) · Systems Practice · doi:10.1007/bf01059830 · Cited by 3,244 · cited 1×
“Similarly, Actor–Network Theory (ANT) [15, 50, 51] conceptualizes technologies as active participants within sociotechnical networks, but it attributes a level of agency to AI systems that cannot exercise consciousness, moral judgment, or accountability [11].”
Research works printed but not linked: 24
- Alsbrook M. Untangling unreliable citations. Georgetown J Legal Ethics. 2024;37:415.
- Barthes R. The death of the author. In: Readings in the theory of religion. Routledge; 2016. p. 141–5.
- Bolukbasi T, Chang K-W, Zou JY, Saligrama V, Kalai AT. Man is to computer programmer as woman is to homemaker? Debiasing word embeddings. In: Advances in neural information processing systems. 2016, vol. 29, p. 4349–4357.
- Cullen CS. Two steps forward or one step back? How restorative justice practices in academic misconduct cases impact student moral decision-making (Publication No. 311) [Doctoral dissertation, University of Georgia]. Electronic Theses and Dissertations. 2025
- Darvin R. The need for critical digital literacies in generative AI-mediated L2 writing. J Second Lang Writ. 2025;67:101186. https://doi.org/10.1016/j.jslw.2024.101186. · doi:10.1016/j.jslw.2024.101186
- Eagleton T. How to read literature. Yale University Press; 2011.
- Elbow P. Writing with power: techniques for mastering the writing process. Oxford University Press; 1983.
- Elsevier. Generative AI policies for journals. 2026. https://www.elsevier.com/about/policies-and-standards/generative-ai-policies-for-journals
- Jurafsky D, Martin JH. Speech and language processing. 3rd ed. Pearson; 2023.
- Khan ZI. Awareness, adoption, challenges and effectiveness of AI tools in research writing: perspectives of research scholars from India and the United Arab Emirates. J Web Libr. 2025;19(3–4):221–55.
- Latour B. On actor-network theory: a few clarifications. Soz Welt. 1996;47(4):369–81.
- Ljubojević D. Student feedback on the effectiveness of ChatGPT for essay writing. In: The fifteenth international conference on e-learning (eLearning-2024). Metropolitan University; 2024, p. 116–126.
- McLuhan M. The medium is the message: an inventory of effects. Bantam Books; 1967.
- Miao F, Holmes W. AI and education: a guidance for policymakers. Unesco Publishing; 2021.
- Mukherjee AN. Application of artificial intelligence: benefits and limitations for human potential and labor-intensive economy–an empirical investigation into pandemic ridden Indian industry. Manag Matters. 2022;19(2):149–66.
- Paul R, Elder L. Critical thinking: the nature of critical and creative thought. J Dev Educ. 2006;30(2):34–5.
- Reyneke Y. Promoting plagiarism prevention in open distance E-learning accounting education: towards an ethics of care conceptual framework (Doctoral dissertation, University of South Africa). 2025
- Sevnarayan K, Potter MA. Generative artificial intelligence in distance education: transformations, challenges, and impact on academic integrity and student voice. J Appl Learn Teach. 2024;7(1):104–14.
- Shahini A. Emotional dimensions of feedback: how AI and human responses shape ESL learning outcomes. Ampersand. 2025;15:100192. https://doi.org/10.1016/j.amper.2025.100192. · doi:10.1016/j.amper.2025.100192
- Sosa E. Epistemic agency. J Philos. 2013;110(11):585–605. https://doi.org/10.5840/jphil2013110110. · doi:10.5840/jphil2013110110
- Springer. Artificial intelligence. 2026. https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500?srsltid=AfmBOopHXGX9iPoXixw8AQXXZbi0NXJpCE8RYZFOIz69RBwrUR2dIm_y
- Trist EL, Emery FE. Sociotechnical systems theory. In: Miner JB, editor. Organizational behavior 2. Routledge; 2015. p. 169–94.
- Wiley. AI guidelines. 2026. https://onlinelibrary.wiley.com/pb-assets/assets/25890514/Generative-AI-Policy-CDTM-1737368544543.pdf
- Zhai X. ChatGPT: a disruptive opportunity for education. TechTrends. 2022;67:76–80. https://doi.org/10.1007/s11528-022-00779-y. · doi:10.1007/s11528-022-00779-y
Metadata
- Authors
- Shahini, A.
- Affiliations
- Islamic Azad University of Gorgan (IR)
- Corresponding author
- Shahini, Amin
- Venue
- Discover Education
- Published
- 2026-08-29
- Volume
- 5
- Issue
- 1
- ISSN
- 2731-5525 (ISSN-L)
- OpenAlex
- W7204631244
- Cited by
- 0
- References
- 57
- FWCI
- 0.00
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by-nc-nd
- creativecommons.org/licenses/by-nc-nd/4.0 (tdm)
- creativecommons.org/licenses/by-nc-nd/4.0 (vor)
- Locations
- Discover Education (publishedVersion, cc-by-nc-nd)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Education › Artificial Intelligence in Education
- Social Sciences › Education › Academic integrity and plagiarism
- Keywords
- AI-assisted research, responsible AI, human-AI collaboration, academic integrity, algorithmic bias, transparency, authorship, intellectual property, AI literacy, institutional governance, higher education, research ethics
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: Developing the HAKI model as a conceptual framework for AI-assisted academic research. https://app.cochat.ai/papers/10.1007/s44217-026-02041-4
- The paper
- Shahini, A. (2026). Developing the HAKI model as a conceptual framework for AI-assisted academic research. Discover Education. https://doi.org/10.1007/s44217-026-02041-4
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.