Pedagogical Applications of Generative AI in Higher Education: A Systematic Review of the Field
Open accessThe DOI matches its Crossref record; 74 of the 75 references it cites were found in a registry or an index; 7 of the 30 sources its claims cite were checked for support.
Abstract
The release of ChatGPT in late 2022 marked the beginning of a rapid transformation in higher education, soon followed by the development of multimodal generative AI programs. As this technology becomes increasingly integrated into teaching and learning, it is crucial to evaluate its current use and impact. This systematic literature review captures the initial academic response to generative AI, providing insights into how higher education has adopted this transformative technology in its first two years. The findings indicate that while some themes from the pre-ChatGPT era persist, new and emerging trends—particularly in fostering creativity, critical thinking, learning autonomy, and prompt literacy—are now taking shape. This shift underscores a growing emphasis on the pedagogical integration of generative AI. However, the review also highlights a key tension: while generative AI enhances efficiency, it raises concerns about overreliance, potentially leading to the outsourcing of critical cognitive and metacognitive skills. To address these challenges and fully harness the potential of generative AI, future research should focus on exploring multimodal generative AI tools and fostering student–teacher-AI collaboration.
Overview
CoChat Summary
TL;DR
This systematic review synthesizes 37 peer-reviewed empirical studies from 2023–2024 on pedagogical uses of generative AI in higher education after ChatGPT. It finds ChatGPT-dominated, text-centric adoption with three pedagogical themes and four challenge areas, highlighting efficiency gains alongside risks of overreliance and underdeveloped multimodal/ethical integration.
Key findings
- From 262 initial records, 37 studies were included; 121 (46%) were excluded because they relied primarily on surveys/interviews about perceptions rather than direct teaching/learning applications. Geographic concentration was China including Hong Kong (11 publications) and the U.S. (8).
- Disciplinary distribution: Education 15 studies; Writing and English Language Learning 13 (Writing/Composition 5, ESL/EFL 8); Computer Science 5; Business 2; Biology 1; Medicine 1. ChatGPT was the dominant tool, with customized GPTs, RAG chatbots, image/video tools, and writing-support tools emerging.
- The 42 pedagogical codes were grouped into three themes: automated feedback and assessment; learning supports; and development of critical skills (creativity, critical thinking, learner autonomy, prompt literacy).
- The 25 challenge codes were grouped into four areas: technical/usability/scalability; quality and ethical concerns; pedagogical challenges; and AI literacy/dependency. A key tension is that GenAI enhances efficiency but raises concerns about overreliance and outsourcing critical cognitive and metacognitive skills.
Methods
- PRISMA-guided systematic review. Searched ERIC, Web of Science, and ScienceDirect using terms for generative AI, higher education, and teaching and learning; included peer-reviewed journal articles published in 2023 and 2024; excluded conference papers, theoretical/conceptual articles, and systematic reviews.
- Coded descriptive metadata and used grounded coding (open, focused, theoretical) plus thematic analysis, producing 42 pedagogical codes and 25 challenge codes. NVivo 14 generated preliminary coding suggestions; a subset of 4 articles (10% of the dataset) was cross-verified with the Coding Comparison Query, reaching 100% consensus after two rounds.
Limitations
- Stated: geographic concentration in China/Hong Kong and the U.S. limits generalizability; excluding 121 survey-based studies (46%) leaves stakeholder attitudes, perceptions, and experiences unexamined; excluding non-peer-reviewed and conceptual/theoretical sources risks omitting emerging innovations or frameworks.
- Evident from the text: a single author conducted manual coding and only a 10% subset was cross-verified with NVivo, limiting inter-rater reliability; the review covers only 37 studies from 2023–2024, so the rapidly evolving GenAI landscape may be underrepresented.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
4 Valid · 5 Not valid · 23 Not checked · 39 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Overall, the current trend suggests a preference for GenAI tools that prioritize simplicity, accessibility, and alignment with core educational tasks, such as writing and critical thinking (Kasneci et al., 2023).”
What the sources say
- Partly supportedcites Kasneci et al., 2023 sourceExperimental — accuracy not yet measured
Our copy supports two of the three elements: Section 2.2 (p. 5) states that 'perceived usefulness, perceived ease of use, and perceived trust in these AI-based tools are determinants to be considered when predicting their acceptance by the teachers' and that perceiving a chatbot as easy-to-use and useful leads to greater acceptance (simplicity), while Sections 1.1-1.2 (pp. 2-3) repeatedly cast writing and critical-thinking development as the main educational applications (alignment with core tasks). However, 'accessibility' appears in our copy only as fair access — multilingual and financial gaps (Section 5, p. 10) — i.e., an equity challenge to be regulated, not a feature users prefer; and the paper nowhere synthesises these into a 'current trend' of user preference (it is a position/review paper, and the acceptance findings it reports are credited to studies [31]-[34]). The triad-as-a-preference-trend framing is therefore not attributable to this source. (LLM, conf=0.68)
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.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“no comprehensive systematic review to date has focused on the pedagogical applications of GenAI in higher education, specifically targeting studies published in 2023 and 2024 - the first two years post ChatGPT.”
What the sources say
- Could not verifycites Zawacki-Richter et al. (2019) sourceExperimental — accuracy not yet measured
The Abstract says this review covers AI applications in higher education in publications from 2007 to 2018. Our copy does not address GenAI reviews of studies published in 2023 or 2024, so it cannot establish whether such a review exists. Searches for ChatGPT, those years, and comprehensive pedagogical reviews found no passage addressing the claim. (LLM, conf=0.77)
- Partly supportedcites Crompton and Burke (2023) sourceExperimental — accuracy not yet measured
The Abstract says this systematic review covers studies from 2016 to 2022, and Gaps and future research says ChatGPT did not appear in the articles it examined and calls for further exploration. That supports a research gap concerning ChatGPT, but the paper does not establish that no comprehensive review covers pedagogical applications of GenAI in studies published in 2023 and 2024. (LLM, conf=0.84)
- Partly supportedcites Crompton and Burke (2024) sourceExperimental — accuracy not yet measured
The Extant Literature section describes earlier education reviews as limited snapshots, with the latest covering publications only through June 2023. The Gaps and Future Research section says this paper examines the first eight months after ChatGPT’s release and focuses on ChatGPT rather than GenAI tools collectively. It does not establish that no comprehensive review of pedagogical GenAI applications in higher education covered studies published in both 2023 and 2024. (LLM, conf=0.82)
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: unclear — It is unclear why the paper says it holds.
- Could not verify
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed to ensure transparency, reproducibility, and reliability in the process of article identification, screening, and selection (Page et al., 2021).”
What the sources say
- Partly supportedcites Page et al., 2021 sourceExperimental — accuracy not yet measured
The paper’s introduction describes PRISMA 2020 as guidance for transparently reporting how studies were identified and selected, and its Discussion says the recommendations can facilitate replication. But the ‘Scope of the guideline’ section explicitly says PRISMA is not intended to guide review conduct or assess methodological quality. It therefore does not establish that following PRISMA ensures reliability in the identification, screening, and selection process. (LLM, conf=0.88)
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
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
“Research on GenAI in education is predominantly concentrated in China (including Hong Kong, which is administered as a Special Administrative Region with a high degree of autonomy), with a combined total of 11 publications highlighting a strong regional focus on GenAIenhanced learning.”
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.
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 U.S. follows with eight publications.”
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.
29 more claims
Create a free account to see all 71 claims and to verify the 3 references still open.
- It may state two separate claims as one. 24
- Probably not a substantial claim. 16
- The quote only partly states this claim. 2
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
“Experimental evidence shows that targeted instructional interventions can recalibrate perceived usefulness and easeof-use of generative AI and reduce over-reliance in assessed tasks (Qian, 2025).”
From Human-in-the-loop assessment with AI: implications for teacher education in Ibero-American universities (2025) · cites this paper as Qian, 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
“Recent higher-education evidence shows that GenAI can support creativity, critical thinking, learning autonomy, prompt literacy, iterative feedback, and rapid variation when embedded within structured pedagogy, but its value declines when tool use is treated as unsupervised production rather than guided learning (Qian 2025).”
From A Study on the Role of Midjourney, A Generative Artificial Intelligence Tool, in Enhancing Student Creativity in Illustration Design Classes (2026) · cites this paper as Qian 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
“This interpretation aligns with designeducation evidence showing that GenAI is most productive when it supports ideation, comparison, and reflection within a guided workflow (Qian 2025).”
From A Study on the Role of Midjourney, A Generative Artificial Intelligence Tool, in Enhancing Student Creativity in Illustration Design Classes (2026) · cites this paper as Qian 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
“Systematic review and Meta-analysis indicate that generative AI may have a positive impact on the motivation and enthusiasm of students studying at higher learning institutions (Xia et al., 2025; Qian, 2025; Chen and Cheung, 2025).”
From Generative AI an academic equalizer? The differential impact of AI-assisted learning on self-efficacy and intrinsic motivation among university students (2026) · cites this paper as Qian, 2025
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“However, GenAI can also undermine learning when students rely on these tools to perform the cognitive work that instructional activities are intended to develop.”
From Generative AI Expands the Intellectual Reach of Course Based Undergraduate Research Experiences (CUREs) (2026) · cites this paper as Qian 2025
5 more claims about this paper
Create a free account to see every claim other papers make about this one.
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
§ Academic Disciplines · p. 7
“fields such as Biology (Collins et al., 2024), Business (Milić et al., 2024; Netland et al. 2025), and Medicine (Hudon et al., 2024) remain underexplored”
§ Identification · p. 3
“Only peer-reviewed journal articles published in 2023 and 2024 were included in this review.”
Citations
Cites
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · Systematic Reviews · doi:10.1186/s13643-021-01626-4 · Cited by 13,874 · cited 1×
“The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed to ensure transparency, reproducibility, and reliability in the process of article identification, screening, and selection (Page et al., 2021).”
Methods for the thematic synthesis of qualitative research in systematic reviews
Thomas, J., & Harden, A. (2008) · BMC Medical Research Methodology · doi:10.1186/1471-2288-8-45 · Cited by 9,312
Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory
Wuetherick, B. (2010) · Canadian Journal of University Continuing Education · doi:10.21225/d5g01t · Cited by 8,311
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/v7gm2_v1 · Cited by 8,090
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) · doi:10.35542/osf.io/5er8f · Cited by 7,552
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
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 2×
“…Science’s representation (Allen et al., 2024; Groothuijsen et al., 2024; Song et al., 2024; Yilmaz and Yilmaz, 2023) aligns with previous findings by Zawacki-Richter et al. (2019) on pre-GenAI programs, reflecting the discipline’s natural synergy with AI capabilities in programming, data analysis,…”
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 · cited 2×
“The National AI Initiative (2021), a U.S. policy framework, advocates embedding AI literacy into curricula, suggesting that courses on prompt design and AI limitations could help mitigate overreliance.”
ELIZA—a computer program for the study of natural language communication between man and machine
Weizenbaum, J. (1966) · Communications of the ACM · doi:10.1145/365153.365168 · Cited by 4,452 · cited 1×
“GenAI’s roots extend back to the 1960 s and 1970 s, during which time early AI-based text generation tools, such as MIT’s ELIZA (Weizenbaum, 1966), SCHOLAR (Car-bonell, 1970), and MYCIN (Shortliffe et al., 1975), were developed.”
Artificial intelligence in higher education: the state of the field
Crompton, H., & Burke, D. (2023) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-023-00392-8 · Cited by 2,010 · cited 1×
“Similarly, Crompton and Burke (2023) reviewed AI research trends between 2016 and 2022 and highlighted a surge in publications on AI applications in education, particularly focusing on language learning and intelligent tutoring systems.”
Research works printed but not linked: 2
- IEEE Spectrum. (2024). What Is Generative AI? Spectrum explains large language models, the transformer architecture, and how it all works. Retrieved February 1, 2025, from https://spectrum.ieee.org/what-is-generative-ai
- Rogers, E. M. (2003). Diffusion of innovations (5th ed.). Free Press.
Cited by
Fajardo-Ramos, D. C., Chiappe, A., & Mella-Norambuena, J. (2025) · Frontiers in Education · doi:10.3389/feduc.2025.1710992 · Cited by 29
“Experimental evidence shows that targeted instructional interventions can recalibrate perceived usefulness and ease-of-use of generative AI and reduce over-reliance in assessed tasks (Qian, 2025).”
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
A systematic review on AI-enhanced pedagogies in higher education in the Global South
Khoza, N. G., & van der Walt, F. (2025) · Frontiers in Education · doi:10.3389/feduc.2025.1667884 · Cited by 47
Shao, H., & Wang, R. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1779741 · Cited by 14
“Mohammed reported enhanced tasks-driven abilities due to the training programme (Mohamed et al., 2024 Areas of Study at the tertiary level (Qian, 2025; Chen and Cheung, 2025).”
Tran, P. T. H., Huynh, L., Bien, T. A., Dang, B., & Nguyen, A. (2025) · Journal of Digital Learning in Teacher Education · doi:10.1080/21532974.2025.2583516 · Cited by 29
“Existing studies have primarily emphasized what GenAI can do for teachers rather than how teachers engage with it strategically and reflectively (Qian, 2025).”
Metadata
- Authors
- Qian, Y.
- Affiliations
- Louisiana State University (US)
- Corresponding author
- Qian, Yufeng
- ORCID
- Qian, Yufeng
- Venue
- TechTrends
- Published
- 2025-06-11
- Volume
- 69
- Issue
- 5
- Pages
- 1105-1120
- ISSN
- 1559-7075 (ISSN-L)
- 8756-3894
- OpenAlex
- W4411166620
- Semantic Scholar
- 280295822be5eb0d8e1cc576f180da10e33721f0
- Cited by
- 206
- References
- 68
- Citations by year
- 2026: 179 · 2025: 27
- FWCI
- 31.38
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- TechTrends (publishedVersion, cc-by)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Education › Educational Technology in Learning
- Keywords
- generative AI, ChatGPT, higher education, pedagogical integration, multimodal generative AI, creativity, critical thinking, learning autonomy, prompt literacy, overreliance on AI, metacognitive skills
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 (qwen3.8-omni-flash and 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: Pedagogical Applications of Generative AI in Higher Education: A Systematic Review of the Field. https://app.cochat.ai/papers/10.1007/s11528-025-01100-1
- The paper
- Qian, Y. (2025). Pedagogical Applications of Generative AI in Higher Education: A Systematic Review of the Field. TechTrends. https://doi.org/10.1007/s11528-025-01100-1
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.