International Journal of Education and Research Vol. 13 No.

9 September 2025

Charting the Rise of ChatGPT in Education: A Bibliometric Analysis of Global Research Trends

Sacide Güzin Mazman Akar

Department of Instructional Technologies, Education Faculty, Usak University. 64200 Usak, Türkiye.
e-mail: [Link]@[Link]
ORCID: [Link]

Abstract
This study provides a comprehensive bibliometric analysis of ChatGPT-related research in education,
mapping its growth trajectory, thematic structures, and geographic distribution since the tool’s public
release in late 2022. By situating ChatGPT scholarship within broader AI-in-education trends, the analysis
identifies dominant publication outlets, most cited works, and critical research gaps. Data were retrieved
from the Web of Science Core Collection for the period 2022–2025, including peer-reviewed journal articles
indexed in SSCI and SCIE. A total of 672 records were analyzed using VOSviewer and bibliometric mapping
techniques to examine publication trends, author collaboration networks, keyword co-occurrence patterns,
and citation structures. Results reveal an exponential increase in publications, from a single article in 2022
to 325 in 2025, with China, the United States, Türkiye, and Australia accounting for over 65% of global
output. Leading publication venues include Education and Information Technologies, BMC Medical
Education, and the Journal of Chemical Education. Thematic analysis identified five major clusters: (1)
technological foundations and large language models, (2) assessment and writing support, (3)
interdisciplinary applications, (4) medical education, and (5) chatbot-supported language learning. Highly
cited works predominantly address academic integrity, student perceptions, and discipline-specific
implementations. However, the field remains methodologically limited, with most studies based on short-
term surveys and experiments.

Keywords: ChatGPT, education, bibliometric analysis, AI in education, research trends, VOSviewer

1. Introduction
In recent years, artificial intelligence (AI) has profoundly influenced education, reshaping how learners
interact with content, instructors, and peers. One of the most transformative developments has been the
rise of generative AI, with OpenAI’s ChatGPT emerging as a particularly impactful tool. Since its public
launch in late 2022, ChatGPT has attracted global attention for its capacity to generate human-like text,
deliver instant feedback, and assist with diverse educational tasks. As noted by recent scholarship, “the
release of ChatGPT marked a turning point in public engagement with AI, combining advanced natural
language processing with unprecedented accessibility” (Deng et al., 2025; von Garrel & Mayer, 2023).
ChatGPT reached over one million users within days of its release, representing one of the fastest adoption
rates for any digital tool in recent history (Grassini, 2023). Recent analyses further emphasize that “the
introduction of ChatGPT has brought significant attention to the educational domain due to its potential to
transform teaching, learning, and assessment practices” (Zhu et al. 2023).
While some praise its ability to enhance engagement, personalize learning, and support formative
assessment, others raise concerns about academic integrity, critical thinking, and the erosion of students'

23
 ISSN: 2411-5681 [Link]

original expression. Educators and policymakers are grappling with how to leverage ChatGPT’s affordances
while mitigating potential risks such as plagiarism, misinformation, and inequitable access (Trust et al.,
2023). Moreover, ChatGPT has been suggested as an academic aide, generating explanations, examples,
and even lesson plans—helping both learners and educators streamline the educational process (Adeshola
& Adepoju, 2023). Generative AI tools like ChatGPT differ from earlier AI applications in education by
enabling “open-ended, conversational interactions that simulate human discourse” (Lo et al., 2024), thus
altering the dynamics of learner–technology interaction. . Its ability to “act as a personalized tutor,
providing immediate feedback and explanations tailored to individual learner needs” further distinguishes it
from previous tools (Zhu et al. 2023).
The growing volume of scholarly work on ChatGPT in educational contexts reflects this duality. Researchers
have explored ChatGPT's role as a cognitive partner, writing assistant, feedback generator, and even a
simulated peer in collaborative learning environments (Teng, 2024; Imran & Almusharraf, 2023). At the
same time, critical voices have highlighted the risks of over-reliance, automation bias, and ethical
ambiguity, particularly in assessment and knowledge construction (Božić et al., 2024; Halaweh, 2023;
Rahman & Watanobe, 2023). Rahman and Watanobe (2023) emphasized that “without explicit guidelines
and AI literacy initiatives, the risk of misuse in assessment and scholarly writing will remain high.” Similarly,
Xiao et al. (2025) stress that “concerns include plagiarism, erosion of academic integrity, and the
outsourcing of cognitive effort to machines.” These concerns align with recent findings noting “the
challenge for educators is to balance the benefits of AI tools with the need to maintain students’
independent problem-solving abilities” (Zhu et al. 2023).
While the scholarly literature on ChatGPT has grown rapidly, prior reviews have often taken a broader AI-in-
education perspective, overlooking ChatGPT-specific patterns across disciplines, education levels, and
geographical contexts. Existing bibliometric analyses have not yet comprehensively mapped thematic
developments, methodological trends, and collaborative networks in this domain. This study addresses that
gap by systematically analyzing publications from 2022 to 2025, with the aim of identifying dominant
research themes, methodological approaches, disciplinary focuses, and collaboration patterns.
 RQ1: How has the number of publications on ChatGPT in education evolved over time between
2022 and 2025?
 RQ2: Which countries have contributed most to the scholarly literature on ChatGPT in educational
contexts?
 RQ3: Who are the most prolific authors in the field?
 RQ4: Which journals publish the most research on ChatGPT in education, and which publications
are most frequently cited?
 RQ5: What are the major research themes and keywords emerging from the literature on ChatGPT
in education?
 RQ6: How have the research themes related to ChatGPT in education evolved over the 2022–

24
 International Journal of Education and Research Vol. 13 No. 9 September 2025

2. Theoretical Background
2.1 Pedagogical Implications of Transition from Artificial Intelligence to Generative AI
AI has evolved from early rule-based systems to advanced machine learning models capable of natural
language processing (NLP) and human-like interaction. Generative AI represents the latest frontier,
producing coherent, contextually relevant text, images, and other media. ChatGPT, based on transformer
architectures (GPT-3.5 and GPT-4), exemplifies this shift, enabling sustained and dynamic exchanges rather
than static outputs (Lo et al., 2024). This transition marks a paradigm shift in educational technology—from
AI as a background computational tool to an active conversational partner in teaching and learning
(Adeshola & Adepoju, 2023; Deng et al., 2025).
ChatGPT functions as more than a tool; it can act as a pedagogical agent capable of influencing the learning
process. Studies report its utility in instructional design, lesson planning, example generation, and
translation of abstract concepts into accessible language (Ngo, 2023; Pokkaliah et al., 2023). Rather than
simply delivering static content, it facilitates interactive and responsive learning, adapting to learners’
needs (Adeshola & Adepoju, 2023).
In pre-service teacher education, ChatGPT supports both content knowledge acquisition and pedagogical
reasoning, embedding itself in the “what” and “how” of teaching (Memarian & Doleck, 2023). Additionally,
it has been used to foster writing, comprehension, and conceptual elaboration (Mosaiyebzadeh et al.,
2023), offering opportunities for scaffolding complex tasks. As Lee & Wu (2025), “educators are beginning
to explore ChatGPT’s potential as a partner in instructional design, formative feedback, and scaffolding
complex tasks.”
However, as Pradana et al. (2023) caution, without critical engagement, such tools may inhibit the
development of higher-order thinking skills. Wei et al. (2025) similarly observes that “when used critically,
ChatGPT can foster higher-order thinking by prompting learners to question, refine, and extend their
ideas,” but this benefit is contingent on intentional pedagogical design.

2.2 Risks and Ethical Challenges

Despite its pedagogical potential, ChatGPT also presents notable challenges, particularly in relation to over-
reliance, academic dishonesty, and the erosion of original thinking (Božić et al., 2024; Halaweh, 2023).
Without explicit guidance, students may over-rely on AI-generated outputs, bypassing opportunities to
engage in critical thinking and independent problem-solving (Trust et al., 2023). A key concern is
automation bias, whereby learners uncritically accept AI-generated outputs, potentially undermining
metacognitive regulation and independent judgment (Lim, 2025; Fan et al., 2024). As Xiao et al. (2025)
observe, these risks encompass plagiarism, the deterioration of academic integrity, and the outsourcing of
cognitive effort to machines.
Moreover, issues of equity and access remain salient, as the benefits of generative AI integration are not
uniformly distributed across educational systems, disciplines, or socioeconomic contexts. Addressing such
disparities requires strategic safeguards to mitigate misuse and ensure that AI adoption aligns with
educational values (Rahman & Watanobe, 2023). In this regard, comprehensive ethical frameworks are
essential—clarifying authorship, delineating intellectual responsibility, and incorporating cultural and
contextual considerations—so that generative AI is conceptualized not merely as a technological tool but as
a transformative force reshaping the epistemological and normative foundations of education. As
emphasized by Von Garrel and Mayer (2023), educational institutions should establish transparent policies

25
 ISSN: 2411-5681 [Link]

for AI use, embed AI literacy training within curricula, and critically evaluate the long-term implications of
generative AI on learning outcomes.
3. Method
3.1 Data Source and Search Strategy
The data for this bibliometric analysis were retrieved from the Web of Science Core Collection (WoSCC), a
database widely recognized for its comprehensive coverage of high-quality scholarly publications. WoSCC
was selected because it indexes journals with established academic credibility in both the Social Sciences
Citation Index (SSCI) and Science Citation Index Expanded (SCIE). Other databases such as Scopus or ERIC
were not included in order to maintain consistency in indexing standards and citation metrics, and to avoid
potential overlaps or discrepancies between databases. While this choice ensures data reliability and
quality, it also introduces a coverage bias, as relevant publications indexed exclusively in other databases
may not be represented.
The search was conducted on [25/07/2025], using the following query:
TS=("ChatGPT" AND "education") OR TS=("ChatGPT" AND "learning") OR TS=("ChatGPT" AND "teaching")
OR TS=("ChatGPT" AND "instruction")
All search terms were enclosed in straight quotation marks to ensure accurate retrieval, avoiding
typographic quotation marks that can cause mismatches in search results. The search was restricted to
publications in the Education Educational Research category, written in English, classified as peer-reviewed
journal articles, and published between 2022 and 2025.
3.2 Inclusion and Exclusion Criteria
The inclusion criteria for the study were as follows:
1. The publication explicitly addressed ChatGPT in the context of education, teaching, learning, or
instruction.
2. The publication was indexed in SSCI or SCIE within the Education Educational Research category.
3. The article was written in English and published between 2022 and 2025.
4. The document type was peer-reviewed journal article.
Exclusion criteria included:
1. Publications not directly related to educational contexts (e.g., technical AI development papers
without an educational focus).
2. Non-article formats such as editorials, letters, conference abstracts, or book reviews.
3. The publication indexed in AHCI or ESCI
4. Duplicates and retracted papers.

3.3 Data Extraction and Preparation

The bibliographic records of all retrieved publications were exported from the Web of Science Core
Collection (WoSCC) in plain text format, including full records and cited references. The exported dataset
was imported into Microsoft Excel for preliminary processing, which involved the removal of duplicate
entries, incomplete records, and studies irrelevant to the scope of this research.

26
 International Journal of Education and Research Vol. 13 No. 9 September 2025

For each included publication, the following information was extracted:

 Bibliographic details (title, authors, publication year, journal, DOI)
 Author affiliations and corresponding countries
 Keywords (both author-provided keywords and Keywords Plus)
 Citation counts
 Abstracts (for subsequent thematic analysis)

During the keyword cleaning process, semantically identical or closely related terms were standardized to
ensure consistency and to prevent fragmentation in the co-occurrence network. This step was crucial for
generating accurate and meaningful clusters in VOSviewer. For example, “AI” was merged with “artificial
intelligence”; “GenAI” and “generative AI” were merged with “generative artificial intelligence”; “chatbot”
was merged with “chatbots”; and “large language model” was merged with “large language models.” Such
standardization reduced redundancy and ensured that conceptually similar keywords were analyzed as
unified entities within the bibliometric mapping.

3.4 Data Analysis

The bibliometric analysis combined descriptive statistics and thematic and temporal mapping. Descriptive
Analysis included annual publication trends, distribution by journals, authors, institutions, and countries,
citation metrics (total citations, average citations per article). Thematic and temporal mapping included
keyword co-occurrence clusters to identify major research themes. The visualizations produced by
VOSviewer were interpreted to identify patterns of term relationships. For the keyword co-occurrence
analysis conducted in VOSviewer, a minimum occurrence threshold of five was applied, and the full
counting method was used to generate the co-occurrence network and thematic clusters.

4. Findings
4.1 Number of publications on ChatGPT in education evolved over time
The distribution of the 664 publications examined in the bibliometric analysis by year indicates that the
research topic first gained visibility in the academic field in 2022. As shown in Figure 1 while only a single
publication (0.15%) appeared in 2022, the number rose sharply to 71 (10.69%) in 2023, 267 (40.21%) in
2024, and 325 (48.95%) in 2025, indicating an exponential growth trend. With the rapid increase in interest,
the number of publications reached 267 (40.21%) in 2024 and 325 (48.95%) in 2025. These findings
demonstrate that ChatGPT became a prominent subject in academic discussions within the educational
context, particularly in 2024 and 2025, reflecting its growing relevance and the continuing upward trend in
research.

27
 ISSN: 2411-5681 [Link]

Figure 1: The distribution of ChatGPT in Education Publication by Year

4.2 Top Contributing Countries in ChatGPT-Education Research

The bibliometric analysis reveals a strong global interest in ChatGPT's application in education, with
contributions spanning over 70 countries. The leading contributors are China (n = 173, 26.05%), the United
States (n = 165, 24.85%), Türkiye (n = 49, 7.38%), and Australia (n = 47, 7.08%), which together account for
over 65% of all publications (Figure 2).

Figure 2. Top 20 Countries with the Number of Publications on ChatGPT in Education

To account for population size, the number of publications per million inhabitants was calculated, revealing
that smaller nations with strong research infrastructures—such as Australia and the United Kingdom—have
28
 International Journal of Education and Research Vol. 13 No. 9 September 2025

a higher relative research output compared to larger countries like China and the United States. This
normalization highlights countries with disproportionately high engagement in ChatGPT-in-education
research despite smaller populations or academic communities.

4.3 Leading Journals and Most Cited Publications in ChatGPT-in-Education Research

According to Web of Science data covering 2022–2025, limited to SSCI and SCIE indexed journal articles in
English, ChatGPT-related educational research is concentrated in several key outlets. The most prolific
journal is Education and Information Technologies (n = 86), followed by BMC Medical Education (n = 53)
and the Journal of Chemical Education (n = 34) (Figure 3). Other notable venues include Medical Teacher (n
= 29), Educational Technology & Society (n = 27), IEEE Transactions on Learning Technologies (n = 24),
International Journal of Educational Technology in Higher Education (n = 24), Interactive Learning
Environments (n = 20), System (n = 18), and the European Journal of Education (n = 18).
These findings show that ChatGPT-in-education research spans discipline-specific (e.g., medical education,
chemistry education) and interdisciplinary technology-enhanced learning journals, with a notable
dominance of technology and health-related venues. Publication trends indicate that most of these articles
appeared in 2023 and 2024, coinciding with the rapid expansion of ChatGPT scholarship.

Figure 3. Top 10 Journals Publishing ChatGPT-in-Education Research (2022–2025)

Citation analysis revealed that the most highly cited works in this domain are predominantly early, high-
impact publications addressing both conceptual and empirical aspects of ChatGPT’s role in education. The
top five papers alone account for over 35% of all citations in the dataset, highlighting a strong
concentration around a small number of seminal works.

29
 ISSN: 2411-5681 [Link]

Table 1: Top 10 Most Cited Publications in ChatGPT-in-Education Research (2022–2025)

Rank Authors Year Title Source Citations
Cotton, D. R. E.; Chatting and cheating: Ensuring
Innovations in Education and
1 Cotton, P. A.; 2023 academic integrity in the era of 781
Teaching International
Shipway, J. R. ChatGPT
International Journal of
Chan, C. K. Y.; Hu, W. Students’ voices on generative
2 2023 Educational Technology in 490
J. AI: Perceptions and experiences
Higher Education
Examining science education in Journal of Science Education
3 Cooper, G 2023 489
ChatGPT: An experimental study and Technology
Lim, W. M.;
Gunasekara, A.; Generative AI applications in International Journal of
4 2023 483
Pallant, J. L.; Pallant, management education Management Education
J. I.
Farrokhnia, M.;
Banihashem, S. K.; A SWOT analysis of ChatGPT: Innovations in Education and
5 2023 468
Noroozi, O.; Wals, A. Implications for education Teaching International
E. J.
A comprehensive AI policy International Journal of
6 2023 education framework for Educational Technology in 351
Chan, CKY university teaching and learning Higher Education
To use or not to use ChatGPT in Interactive Learning
higher education? A study of Environments
7 2024 308
students' acceptance and use of
Strzelecki, A technology
Impact of ChatGPT on learners Education and Information
8 2023 in a L2 writing practicum: An Technologies 262
Yan, D exploratory investigation
Large language models in Education and Information
education: A focus on the Technologies
9 2023 complementary relationship 251
between human teachers and
Jeon, J; Lee, SY ChatGPT
The impact of Generative AI Interactive Learning
(GenAI) on practices, policies Environments
10 Chiu, TKF 203 and research direction in 247
education: a case of ChatGPT
and Midjourney

30
 International Journal of Education and Research Vol. 13 No. 9 September 2025

The most cited article is by Cotton, Cotton, and Shipway (2023), which examines ChatGPT use in higher
education through the lens of academic integrity, with 781 citations. This is followed by Chan and Hu
(2023), who explore students’ perceptions and experiences with generative AI (490 citations). An
experimental study on ChatGPT’s potential in science education ranks third (489 citations), while a paper on
generative AI in management education is fourth (483 citations). The fifth, by Farrokhnia, Banihashem,
Noroozi, and Wals (2023), presents a SWOT analysis of ChatGPT in educational contexts (468 citations).
These findings indicate that the most cited literature on ChatGPT in education predominantly focuses on (i)
ethics, academic integrity, and policy, (ii) student perceptions and experiences, and (iii) discipline-specific
implementations. Furthermore, the citation distribution shows a strong concentration around a small
number of seminal works, suggesting that the field remains anchored to a limited set of guiding references
that continue to influence subsequent research.

4.4 Most Prolific Authors

The bibliometric analysis identified the most productive authors contributing to the scholarly literature on
ChatGPT in educational contexts (Figure 4). The most prolific author is Lee, H.Y., with six publications,
followed by Li, Y., Huang, Y.M., Wu, T.T., Guo, K., Ji, Y., Hwang, G.J., Wang, L., Masters, K., and Shin, D., each
with five publications. While the distribution of publication counts among these authors is relatively close,
Lee, H.Y.’s slight lead suggests an emerging leadership role in shaping the discourse on ChatGPT in
education. As illustrated in Figure 4, the publication distribution among the top authors reflects a relatively
balanced yet influential core of researchers actively engaging in this emerging research domain.

Figure 4: Top 10 prolific authors on ChatGPT in educational contexts.

The concentration of publications among a limited group of researchers indicates the presence of a core set
of authors driving the field forward. This core group likely plays a significant role in setting research
agendas, establishing conceptual frameworks, and fostering collaboration networks in the area of AI-
enhanced education.

31
 ISSN: 2411-5681 [Link]

4.5 The major research themes and keywords emerging from the literature on ChatGPT in education

The co-occurrence network analysis and the cluster statistics collectively reveal five prominent thematic
clusters in the literature on ChatGPT in education. Each cluster is distinguished by its core keywords, cluster
size, and total link strength, reflecting the thematic breadth and relational density of the research
landscape (Table 2).

Table 2: Clusters and representative keywords from the ChatGPT-in-education literature

Cluster (Color in Cluster Total Link
Core Keywords Thematic Focus
Figure) Size Strength
artificial intelligence, large 60 485
Technological foundations
language models, natural
and user acceptance in
Cluster 1 (Red) language processing,
higher education AI
technology acceptance,
adoption
university students,
Assessment, general public, 24 69 Assessment and writing
writing, first-year skills development across
Cluster 2
undergraduate, upper-division academic levels in online
(Green)
undergraduate, curriculum, learning contexts
internet/web-based learning
Chatgpt, generative artificial 56 283 Interdisciplinary
intelligence, higher education, applications of ChatGPT in
Cluster 3 (Blue) engineering education, higher education with a
technology acceptance, focus on engineering and
Colloborative Learning collaborative learning
medical education, critical 16 89 AI-enhanced learning and
thinking, educational critical thinking in medical
Cluster 4
technology, medical students, education
(Yellow)
machine learning, self-directed
learning
15 39 Chatbot-supported writing
Chatbots, engineering
Cluster 5 and language learning
education, L2 writing,
(Orange) integration in engineering
automated writing evaluation
education
Note. Cluster colors correspond to the VOSviewer network map (see Figure 1).
Cluster 1 (Red) is the largest, comprising 60 keywords with a total link strength of 485, and centers on
artificial intelligence, large language models, natural language processing, technology acceptance, and
university students. This cluster primarily reflects research on the technological foundations of ChatGPT
and its adoption within higher education contexts.
Cluster 2 (Green) consists of 24 keywords (total link strength: 69), anchored around assessment, writing,
first-year undergraduate, upper-division undergraduate, curriculum, and internet/web-based learning. The

32
 International Journal of Education and Research Vol. 13 No. 9 September 2025

thematic focus here is on evaluation and writing skills development in online learning contexts across
various academic levels.
Cluster 3 (Blue), with 56 keywords and a total link strength of 283, revolves around ChatGPT, generative
artificial intelligence, higher education, engineering education, technology acceptance, and collaborative
learning. This cluster reflects interdisciplinary applications of ChatGPT in higher education, particularly in
engineering and collaborative learning environments.
Cluster 4 (Yellow) comprises 16 keywords (total link strength: 89) such as medical education, critical
thinking, educational technology, medical students, machine learning, and self-directed learning. It
represents studies integrating AI tools in medical training, often with a focus on fostering critical thinking
skills.
Cluster 5 (Orange), the smallest with 15 keywords and a total link strength of 39, includes chatbots,
engineering education, L2 writing, and automated writing evaluation. This cluster captures research on
chatbot-supported applications for writing and language learning, particularly in engineering education
contexts.

Figure 5: Keyword co-occurrence network of ChatGPT-in-education literature (VOSviewer output)

The network visualization (Figure 5) illustrates how these clusters interconnect, with high-density linkages
between the artificial intelligence and ChatGPT nodes in the red and blue clusters, indicating their centrality
in the research discourse. Similarly, the green and orange clusters share thematic overlaps in writing and
assessment, while the yellow cluster maintains more discipline-specific connections in medical education.
Together, these findings highlight both the thematic diversity and the interdisciplinary nature of ChatGPT-
related educational research.

Overall, the network structure reveals that research on ChatGPT in education is multi-clustered, bridging
technical AI research with practical educational implementation and ethical debates. The strong
33
 ISSN: 2411-5681 [Link]

interconnections between clusters indicate that most studies adopt interdisciplinary perspectives,
combining pedagogy, assessment, ethics, and AI technology.

5. Discussion
This bibliometric analysis provides a comprehensive overview of the emerging body of literature on
ChatGPT in education, situating its development within broader AI-in-education research trends. By
integrating bibliometric findings with insights from recent empirical and conceptual studies (e.g., Guo et al.,
2024; Pradana et al., 2023; Grassini, 2023; Lee, 2024), the discussion highlights temporal growth patterns,
geographic and disciplinary distributions, thematic structures, and critical research gaps.
The rapid escalation in ChatGPT-related educational research following its public release in late 2022
reflects the tool’s exceptional global uptake. While only a single publication (0.15%) appeared in 2022, the
number rose sharply to 71 (10.69%) in 2023, 267 (40.21%) in 2024, and 325 (48.95%) in 2025. This
trajectory mirrors patterns observed in broader AI-in-education research, where initial feasibility
discussions quickly evolve into more diverse applications (Guechairi, 2024; Guo et al., 2024). Consistent
with Lo’s (2023) rapid review, this surge reflects both the novelty of the technology and its perceived
capacity to address persistent educational challenges such as scalable individualized support, efficient
feedback mechanisms, and flexible knowledge access. However, as both Lo (2023) and Zhu et al. (2023)
emphasize, the majority of current research is exploratory in nature, with limited longitudinal evidence to
confirm sustained educational benefits.
This growth has been accompanied by a clear geographic concentration of research activity. Publications
are heavily clustered in China, the United States, Türkiye, and Australia, which together account for more
than 65% of the total. Such dominance underscores both the global appeal of ChatGPT and the persistence
of digital research divides, with contributions from regions such as Africa and parts of South America
remaining minimal. Similar imbalances have been reported in related bibliometric reviews (Akhmadieva et
al., 2023), suggesting that infrastructural capacity, funding availability, and policy frameworks play decisive
roles in determining research participation. Lo (2023) further notes that regional adoption is often
influenced by language accessibility, regulatory stance toward AI, and institutional readiness, factors that
merit closer examination in future studies. This geographic concentration is mirrored in the authorship
landscape, where a relatively small core group of scholars—often working within medical education,
computer science education, and applied linguistics—drive the field. Although interdisciplinary engagement
is present, collaborations often occur within parallel but partially overlapping networks, limiting broader
cross-disciplinary integration.
The journal analysis further illustrates how publication activity is concentrated in specific domains.
Technology-enhanced learning and medical education journals dominate the field, with Education and
Information Technologies, BMC Medical Education, and the Journal of Chemical Education emerging as
leading venues. The most frequently cited works tend to be systematic reviews and conceptual syntheses,
reflecting a scholarly demand for theoretical consolidation in a rapidly developing field. These findings align
with the prominence of thematic clusters identified in the keyword co-occurrence analysis, which span
technological foundations, assessment and writing, interdisciplinary applications, medical education, and
chatbot-supported language learning. The strong interconnections between these clusters indicate a
research landscape in which technological affordances are being explored alongside subject-specific
pedagogical strategies.

34
 International Journal of Education and Research Vol. 13 No. 9 September 2025

The reviewed literature positions ChatGPT as a technology with substantial potential to transform
educational practices, yet its adoption is accompanied by a combination of opportunities, challenges, and
unresolved research gaps. Teacher education programs must equip future educators with the AI literacy
skills needed to critically evaluate and effectively integrate tools like ChatGPT in ways that enhance, rather
than undermine, student learning (Trust et al., 2023). Pedagogically, ChatGPT offers notable affordances
such as personalized learning experiences, automated formative feedback, writing assistance, and
multilingual support. Studies indicate its capacity to enhance idea generation, facilitate collaborative
problem-solving, and support domain-specific learning, particularly in STEM and language education
(Grassini, 2023; Wei et al., 2025). Furthermore, ChatGPT has been found to improve accessibility for
learners with different needs, foster motivation, and support self-paced learning when appropriately
integrated into instructional design (Oranga, 2023). This echoes Zhu et al.’s (2023) findings that educators
have begun to harness ChatGPT to provide differentiated instruction and scaffold complex cognitive tasks,
yet many implementations remain.. However, these benefits are primarily documented in short-term
interventions, leaving questions about their long-term impact unanswered.
Alongside these opportunities, ethical and academic integrity concerns remain prominent. Risks such as
plagiarism, the erosion of original thinking, and automation bias—where learners accept AI-generated
outputs without critical evaluation—are consistently identified in the literature (Xiao et al., 2025; Rahman
& Watanobe, 2023). While some institutions have begun developing guidelines and AI literacy initiatives to
mitigate these risks, the extent to which such measures are effective across different educational and
cultural contexts remains largely unexplored. Addressing these challenges will require a more deliberate
integration of governance and policy frameworks that emphasize transparent institutional guidelines,
targeted AI literacy training, and culturally responsive approaches (Von Garrel & Mayer, 2023).
Methodologically, the current body of research is dominated by perception-based surveys and short-term
experiments. While these approaches are valuable for capturing early adoption patterns, they are
insufficient for understanding longitudinal outcomes or variations across cultural and disciplinary contexts.
The field would benefit from mixed-method designs, comparative cross-disciplinary experiments, and long-
term tracking of learning outcomes. Moreover, the observed geographic disparities highlight the need for
targeted efforts to support research in underrepresented regions, addressing barriers such as limited
infrastructure, insufficient teacher training, and lack of supportive policy environments.
In summary, the thematic structure identified in this study reflects a field in transition—from early
explorations of technological capacity toward more sophisticated pedagogical applications and ethical
considerations. The strong interconnections between thematic clusters indicate that effective integration of
ChatGPT in education will depend not only on technological advancement but also on deliberate
instructional design, equity-focused policy measures, and evidence-based governance. Bridging the gap
between conceptual promise and long-term educational impact will require sustained empirical inquiry,
interdisciplinary collaboration, and the inclusion of currently underrepresented voices in the global
research landscape.

6. Conclusion
This study makes a distinctive contribution to the rapidly expanding body of literature on ChatGPT in
education by offering the first comprehensive bibliometric mapping of publications between 2022 and 2025
indexed in SSCI and SCIE under the Education Educational Research category. Theoretically, it advances

35
 ISSN: 2411-5681 [Link]

understanding by identifying five interconnected thematic clusters that reveal how pedagogical,
technological, and ethical dimensions converge in the emerging research landscape. Methodologically, it
applies a rigorous, transparent bibliometric approach using WoSCC data, standardized keyword cleaning,
and VOSviewer-based co-occurrence analysis, enabling both descriptive and thematic insights. Practically,
the findings provide educators, policymakers, and technology developers with an evidence-based overview
of global research trends, prolific authors, influential journals, and emerging themes, informing strategic
integration of ChatGPT into diverse educational contexts. By combining these contributions, the study not
only synthesizes the current state of research but also establishes a foundation for future inquiry that
bridges conceptual understanding and applied implementation.

7. Limitations and Directions for Future Research

While the study offers robust insights, several limitations should be acknowledged. First, the dataset was
restricted to the Web of Science Core Collection, which, although ensuring high-quality indexing, may omit
relevant publications indexed in other databases such as Scopus or ERIC. Second, the inclusion criteria
limited the analysis to English-language journal articles, potentially underrepresenting non-English
scholarship and alternative publication formats. Third, the keyword co-occurrence analysis, while
systematically conducted, relied on a minimum occurrence threshold that may have excluded emerging
niche topics with lower frequency. Finally, the study provides a snapshot up to mid-2025, and the rapidly
evolving nature of ChatGPT-related research means that trends may shift significantly in the near future.

Future research should address these limitations by conducting cross-database bibliometric analyses to
expand coverage and ensure inclusivity of diverse languages and publication types. Longitudinal studies
that track thematic evolution over extended periods are needed to capture the maturation of the field.
Moreover, comparative analyses across cultural, linguistic, and socioeconomic contexts would deepen
understanding of geographic disparities and inform equitable adoption strategies. Integrating bibliometric
mapping with systematic literature reviews or meta-analyses could also enrich interpretation, linking
publication patterns to empirical evidence of pedagogical impact.

Statements and Declarations

Acknowledgments: This research did not receive any specific grants from funding agencies in the public,
commercial, or not-for-profit sectors.

Declarations of interest: The authors declare that they have no conflicts of interest related to this study.

36
 International Journal of Education and Research Vol. 13 No. 9 September 2025

REFERENCES
Adeshola, I., & Adepoju, A. P. (2024). The opportunities and challenges of ChatGPT in education. Interactive
Learning Environments, 32(10), 6159-6172.
Akhmadieva, R. S., Udina, N. N., Kosheleva, Y. P., Zhdanov, S. P., Timofeeva, M. O., & Budkevich, R. L.
(2023). Artificial intelligence in science education: A bibliometric review. Contemporary Educational
Technology, 15(4), ep460. [Link]
Božić, V., & Poola, I. (2023). Chat GPT and education. Preprint, 10.
Deng, R., Jiang, M., Yu, X., Lu, Y., & Liu, S. (2025). Does ChatGPT enhance student learning? A systematic
review and meta-analysis of experimental studies. Computers & Education, 227, 105224.
Fan, Y., Tang, L., Le, H., Shen, K., Tan, S., Zhao, Y., ... & Gašević, D. (2025). Beware of metacognitive
laziness: Effects of generative artificial intelligence on learning motivation, processes, and
performance. British Journal of Educational Technology, 56(2), 489-530.
Grassini, S. (2023). Shaping the future of education: Exploring the potential and consequences of AI and
ChatGPT in educational settings. Education sciences, 13(7), 692.
Guo, S., Zheng, Y., & Zhai, X. (2024). Artificial intelligence in education research during 2013–2023: A review
based on bibliometric analysis. Education and information technologies, 29(13), 16387-16409.
Guechair, S. (2024). Artificial Intelligence in Education: A Comprehensive Bibliometric Study on Scopus
(2010-2024). ATRAS journal, 5(3), 445-463.
Halaweh, M. (2023). ChatGPT in education: Strategies for responsible implementation. Contemporary
educational technology, 15(2).
Imran, M., & Almusharraf, N. (2023). Analyzing the role of ChatGPT as a writing assistant at higher
education level: A systematic review of the literature. Contemporary Educational Technology, 15(4),
ep464.
Lee, H. (2024). The rise of ChatGPT: Exploring its potential in medical education. Anatomical sciences
education, 17(5), 926-931.
Lee, H. Y., & Wu, T. T. (2025). Enhancing Blended Learning Discussions with a Scaffolded Knowledge
Integration–Based ChatGPT Mobile Instant Messaging System. Computers & Education, 105375.
Lim, C. (2025). DeBiasMe: De-biasing Human-AI Interactions with Metacognitive AIED (AI in Education)
Interventions. arXiv preprint arXiv:2504.16770.
Lo, C. K., Hew, K. F., & Jong, M. S. Y. (2024). The influence of ChatGPT on student engagement: A systematic
review and future research agenda. Computers & Education, 219, 105100.
Lo, C. K. (2023). What is the impact of ChatGPT on education? A rapid review of the literature. Education
sciences, 13(4), 410.
Memarian, B., & Doleck, T. (2023). ChatGPT in education: Methods, potentials, and limitations. Computers
in Human Behavior: Artificial Humans, 1(2), 100022.
Montenegro-Rueda, M., Fernández-Cerero, J., Fernández-Batanero, J. M., & López-Meneses, E. (2023).
Impact of the implementation of ChatGPT in education: A systematic review. Computers, 12(8), 153.
Mosaiyebzadeh, F., Pouriyeh, S., Parizi, R., Dehbozorgi, N., Dorodchi, M., & Macêdo Batista, D. (2023,
October). Exploring the role of ChatGPT in education: Applications and challenges. In Proceedings of
the 24th annual conference on information technology education (pp. 84-89).
Ngo, T. T. A. (2023). The perception by university students of the use of ChatGPT in education. International
Journal of Emerging Technologies in Learning (Online), 18(17), 4.

37
 ISSN: 2411-5681 [Link]

Oranga, J. (2023). Benefits of artificial intelligence (ChatGPT) in education and learning: Is Chat GPT
helpful. International Review of Practical Innovation, Technology and Green Energy (IRPITAGE), 3(3),
46-50.
Pokkakillath, S., & Suleri, J. (2023). ChatGPT and its impact on education. Research in Hospitality
management, 13(1), 31-34.
Pradana, M., Elisa, H. P., & Syarifuddin, S. (2023). Discussing ChatGPT in education: A literature review and
bibliometric analysis. Cogent Education, 10(2), 2243134.
Rahman, M. M., & Watanobe, Y. (2023). ChatGPT for education and research: Opportunities, threats, and
strategies. Applied sciences, 13(9), 5783.
Teng, M. F. (2024). “ChatGPT is the companion, not enemies”: EFL learners’ perceptions and experiences in
using ChatGPT for feedback in writing. Computers and Education: Artificial Intelligence, 7, 100270.
Trust, T., Whalen, J., & Mouza, C. (2023). Editorial: ChatGPT: Challenges, opportunities, and implications for
teacher education. Contemporary Issues in Technology and Teacher Education, 23(1), 1-23.
Xiao, Y., Li, D., & Guo, K. (2025). Using ChatGPT to bring non-player characters to life: Effects on students’
storyline-driven game-based writing learning. Computers & Education, 105414.
Von Garrel, J., & Mayer, J. (2023). Artificial Intelligence in studies—use of ChatGPT and AI-based tools
among students in Germany. Humanities and social sciences communications, 10(1), 1-9.
Wang, J., & Fan, W. (2025). The effect of ChatGPT on students’ learning performance, learning perception,
and higher-order thinking: insights from a meta-analysis. Humanities and Social Sciences
Communications, 12(1), 1-21.
Wei, X., Wang, L., Koszalka, T. A., Lee, L. K., & Liu, R. (2025). Enhancing pre-service teachers' reflective
thinking skills through generative AI-assisted digital storytelling creation: A three-dimensional
framework analysis. Computers & Education, 105356.
Zhu, C., Sun, M., Luo, J., Li, T., & Wang, M. (2023). How to Harness the Potential of ChatGPT in
Education?. Knowledge Management & E-Learning, 15(2), 133-152.

38