Mapping the Rise of Generative AI in Personalized Learning: A Global Bibliometric Analysis in Higher Education

Mapping the Rise of Generative AI in Personalized Learning: A Global Bibliometric Analysis in Higher Education | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Mapping the Rise of Generative AI in Personalized Learning: A Global Bibliometric Analysis in Higher Education Nora Azima Noordin This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9113918/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The rapid development of Generative Artificial Intelligence (GenAI), especially large language models, has heightened academic attention on its role in incentivizing personalized learning in higher education. The current paper will include a thematic bibliometric survey of the Scopus-indexed articles published by 2023 and 2025 to trace the changing research environment at the intersection of generative AI and personalized learning. Through the Bibliometrix R package through Biblioshiny and VOSviewer, the analysis is performed to analyze the development of publications, journals and documents with the greatest impact, patterns of international collaboration, and the development of research topics. The results demonstrate a dramatic rise in the number of research and a great involvement of the world, and also the change in the scholarly perspective of research towards pedagogic, ethical and institutional issues. Such themes of research are AI-facilitated personalized learning, student engagement, academic honesty, and responsible usage of AI. This study can inform researchers, educators, and policymakers in the timely manner by synthesizing the global research trends, so that they can design, govern, and assess generative AI-enhanced personalized learning in higher education. Artificial Intelligence and Machine Learning Generative artificial intelligence Personalized learning Higher education Bibliometric analysis Large language models Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction The swift advancement progress of Generative Artificial Intelligence (GenAI), particularly large language models (LLM) like ChatGPT, has brought a complete paradigm shift when it comes to the way digital technologies are being conceptualized and used in the context of higher education. In contrast to earlier types of educational artificial intelligence, which were mostly limited to predictive analytics, rule-based systems or adaptive paths, generative AI systems have the ability to generate human like text, descriptions, responses and teaching materials in real-time. This is an ability that has made GenAI potentially an extraordinary power in the teaching, learning, and assessment processes in institutions of higher learning (Dwivedi et al., 2023 ; Mhlanga, 2023 ). Personalized learning has been known to be a very important pedagogical strategy to deal with the diversity among learners to increase the engagement and also help them to make learning self-regulated in higher learning. Personalization has traditionally been supported using learning analytics, intelligent tutoring systems, and adaptive learning platforms, as well as changing the content or pacing relative to the predefined rules or a learner (Zawacki-Richter et al., 2019 ). Generative AI is a qualitative extension of this paradigm since these systems are dynamically capable of producing customized explanations, formative feedback, and learning materials that shall be in accordance with the needs of the individual learner. According to recent research, GenAI-led personalization can potentially facilitate learner agency and motivation, as well as transform the design and the role of an educator (Huang et al., 2023 ). The interest in generative AI in the academic setting has intensely increased since the open release and popularization of large language models in 2023. A broad variety of uses, such as AI-assisted feedback, automated evaluation, academic writing assistance, tutoring, and learning analytics-based personalization have been discussed by researchers. Meanwhile, this fast spread has brought serious issues regarding the academic integrity, ethically sound application, transparency, and the future implications of generative AI on learning and educational principles (Tang and Xiao, 2023 ; Aljanabi and Noorbehbahani, 2023 ). These two narratives, namely, innovation and risk, have informed the current discourse of GenAI in higher education. Despite this growing body of research, a few review articles and concept papers have discussed artificial intelligence in education more generally, many of them date back to a time when generative AI was at early stages of development, or AI is understood as a uniform group regardless of the type of model (Mustafa et al., 2024 ). As a result, systematic knowledge regarding the development of research at the intersection of generative AI and personalized learning in higher education is limited over the past years, such as the post-LLM emergence period. The bibliometric analysis is an objective and strict method to fill this gap by estimating the pattern of publications, the sources of influence, the network of collaborations, and thematic development in a field of research. Unlike narrative or systematic reviews, a bibliometrical technique can be used to determine macro-level trends and scholarly output structures, which are built upon large quantities of scholarly output (Cobo et al., 2011 ). The use of bibliometric methods on the fast growing literature on generative AI and personalized learning is especially useful in capturing general trends in research and inform future investigation. Accordingly, this study performs a thorough bibliometric search of publications Scopus-indexed between 2023 and 2025 that discuss the topic of generative AI and personalized learning in post-secondary education. The study was examined through the Bibliometrix R package (through Biblioshiny) and VOSviewer which analyzes the growth of publications, leading journals and documents, patterns of collaboration in the world, and development of themes. A modified PRISMA protocol is used to make the data selection and processing transparent and replicable (Page et al., 2021 ). This study makes three key contributions. First, it offers a current and targeted mapping of the research at the interface of generative AI and personalized learning in a crucial time of technological transformations in higher education. Second, it determines prevailing and new research topics, showing a change in the initial technical orientation to the pedagogical, ethical, and institutional issues. Third, it provides researchers, educators, and policymakers with evidence-based insights to design, govern, and evaluate higher education generative AI-enabled personalized learning. This bibliometric review contributes to the current knowledge on how AI generated (generative AI) is transforming personalized learning and facilitates the creation of theoretically, empirically and ethically responsible and pedagogically significant AI uses in higher education. Literature Review Artificial intelligence has been increasingly embedded in higher education through applications such as learning analytics, intelligent tutoring systems, and adaptive learning platforms. The previous systems were majorly based on predictive models and predetermined instructional rules to facilitate learning. Generative AI, and especially large language models (LLMs) like ChatGPT, is a notable rejection of such methods, as these systems are capable of producing natural language output, explanations, and feedback, which are reminiscent of human interaction (Yusuf et al., 2024 ). According to the recent study, the spread of generative AI tools in the higher education sector is going viral and is used in academic writing assistance, tutoring, feedback, and assessment creation (Xia et al., 2024 ). Researchers highlight that generative AI systems cannot be seen as productivity tools only but can be considered a new category of educational technology that can influence learner cognition, engagement, and instructional practices. Simultaneously, the issues of academic integrity, excessive dependence on AIs, and the loss of critical thinking have been given the center stage (Shahzad et al., 2024 ). The literature increasingly conceptualizes generative AI as a pedagogical agent rather than a passive tool. This has led to the demand of new instructional designs, assessment designs, and faculty development plans, that aligns AI utilization with the learning outcomes and educational values (Chan et al., 2023). Consequently, studies on generative AI in higher education have grown at an alarming rate, though conceptually varied and experimental and methodological in nature. Personalized Learning and AI-Driven Adaptation Personalized learning is not a new idea that has been promoted as an effective learning strategy to meet the diversity of learners and enhance learning outcomes in higher education. Learning analytics, adaptive learning systems, and intelligent tutoring systems have been used to support traditional methods of personalization, which varies content and pacing depending on learner data (Fortuna et al., 2025 ). Such systems are, nevertheless, usually based on decision rules which are established and restricted interaction modalities. Generative AI provides a novel aspect of customized learning by offering real-time and conversational, contextual modification. Empirical research proposes that AI-based learning assistants can personalize explanations and create practice resources as well as formative feedback to different needs and preferences of individual learners (Yang and Ogata, 2024 ; Holmes et al., 2022 ). These capabilities can be heard by constructivist and learner-centered pedagogies that focus on active learning and self-regulation. However, another problem that may arise as a result of AI-based personalization is noticeable in the literature as well. The existence of many issues, including algorithmic bias, unequal quality of personalization, and transparency of AI decision-making, casts their doubts on equity and pedagogical validity (Merino-Campos, 2025 ; Vieriu and Petrea, 2025 ). Therefore, researchers put forward that personalization made possible by generative AI should be informed by teaching principles and backed with empirical data about the effectiveness of the learning process as opposed to technological novelty alone (Noordin et al., 2024 ). Ethical, Pedagogical, and Institutional Considerations The fast implementation of generative AI in higher education has heightened ethical, pedagogical, and institutional controversies. Issues related to ethics that are discussed in the literature are the privacy of data, responsibility, bias in the work of AI, and how the use of AI can affect academic integrity (Vorobyeva et al., 2025 ). In assessment contexts, generative AI challenges traditional notions of authorship and originality, prompting reconsideration of assessment design and evaluation criteria. Pedagogically, scholars believe in applying generative AI to curricula in a manner that facilitates AI literacy, critical analysis, and responsible application. Instead of limiting the use of AI tools, most scholars recommend teaching methods that directly involve students in thinking about the advantages and weaknesses of the content created by AI. At the institutional level, universities are coming up with policies and governance structures to control the use of AI in teaching and learning. These policies are region-specific and they capture the difference in regulation environment, technological preparedness and the culture that embraces AI (OECD, 2023 ). Based on the literature, it is stressed that effective and responsible deployment of generative AI needs institutional incentive, faculty preparation, and consistency between technological framework and instructional objectives. Bibliometric Studies on AI and Education The use of bibliometric studies has been important in indexing the research trends of artificial intelligence in education. The present research analyzed the publication growth, collaboration networks, and thematic structures in the fields of learning analytics, intelligent tutoring systems, and educational data mining (Durak et al., 2024 ; Ateş et al., 2025 ). These papers are a rich source of macro-level data on the development of the research on AI. Nonetheless, the majority of available bibliometric reviews are not dated after the popularization of generative AI or are simply conceptualizing AI in general and vague terms. Consequently, they fail to record the unique research dynamics related to large language models or their presence in personalized learning conditions. Few recent reviews consider generative AI explicitly and none of them specifically address higher education settings or thematic development in the post-2023 period. Research Gap and Study Positioning According to the literature that has been reviewed, three major gaps are apparent. To start with, there is very limited recent bibliometric review that specifically addresses the topic of generative AI and personalized learning in higher education after the advent of large language models. Second, the literature supports minimal research to determine the way research themes have changed following the ethical, pedagogical, and institutional issues related to generative AI. Third, patterns of collaboration and contributions to research in this developing area across the globe are underexplored. To address these gaps, the present study applies a narrow bibliometric search of the Scopus-indexed papers published in 2023–2025. This analysis combines performance analysis, collaboration mapping, key word co-occurrence and thematic evolution analysis to have a holistic view of the intellectual, social and conceptual frameworks that define research on generative AI and personalized learning in higher education. By doing so, it builds on previous bibliometric efforts and provides current information on what needs to be done to further the development of theory, research, and practice in educational technology. Methodology This study follows a bibliometric research design, which will identify and synthesize the body of academic literature on the topic of generative AI and personalized learning in higher learning. Bibliometric analysis makes it possible to analyze quantitatively the patterns of publications, the intellectual organization, collaboration patterns, and the development of themes in a research sphere. In order to improve the transparency and replicability of the study, the study adheres to an adapted PRISMA 2020 procedure that is progressively used in bibliometric reviews to report data collection, screening, and inclusion processes. Data Source and Search Strategy The reason to choose Scopus as the main source of data is its extensive coverage of the peer-reviewed journals and conference proceedings in the field of education, technology, and interdisciplinary research. Scopus is a well-known and trustworthy source of bibliometric research and can be easily used with bibliometric analysis software (Bibliometrix and VOSviewer). The search strategy was created to include a literature that covers the combination of generative AI, personalized learning, and higher education. The desired comprehensive search string was built on the base of the key words in the sphere of generative AI technologies (e.g., “generative AI,” “large language models,” “ChatGPT,” “AI chatbots”), personalized learning concepts (e.g., “personalized learning,” “adaptive learning,” “intelligent tutoring systems”), and higher education contexts (e.g., “higher education,” “university,” “tertiary education”). The search was applied to titles, abstracts, and keywords. In order to capture the latest trends, after the massive use of large language models, the search was restricted to articles published in 2023–2025. The last search took place in Scopus in November 2025 and was consistent in all of the searched records. Inclusion and Exclusion Criteria In order to make sure that the dataset remains relevant and of high quality, there were explicit inclusion and exclusion criteria. The criteria of inclusion were peer-reviewed journal articles and conference proceedings, publications in English, and studies that specifically covered the topic of generative AI and personalized or adaptive learning in a higher education context. Publications were excluded if they editorials, short notes, reviews of books, non-academic documents and those published before 2023 were excluded. Upon the use of these criteria and elimination of duplicate records, a final dataset of 282 publications was left to be analyzed. PRISMA-Based Data Screening and Selection The screening and selection process were documented to an adapted PRISMA 2020. First, 299 records were obtained in Scopus according to the set search string and period of time. Redundant records were also detected and eliminated to make 282 distinct publications. Since the bibliometric analysis is a qualitative study, full-text screening was not carried out but rather relevancy was verified by metadata validation, keywords matching and source checking. The PRISMA flow diagram is adapted, which demonstrates every step of the data selection and cleaning process, making it more transparent in the methodology. Data Cleaning and Pre-Processing Before the analysis, a systematic cleaning and pre-processing stage was applied to the bibliographic data to give it accuracy and consistency. To ensure that the dataset uploaded to Scopus contains all the necessary metadata fields, the dataset was checked to ensure the coverage of the following fields authorship, publication year, source title, affiliations, keywords, and abstract. Redundant records were eliminated and inconsistency in author names, affiliation and special characters were regularized to prevent fragmentation in network analyses. Preliminary preprocessing was done through the use of spreadsheet software to look at and rectify the clear formatting errors. The filtered data was subsequently inputted into the R studio software to be prepared further with the help of the Bibliometrix package. Bibliographic areas were all standardized to make them compatible with bibliometric indicators and network studies. Such stage of preprocessing made sure that the final dataset can be analysed effectively in terms of robust performance, collaboration mapping and thematic exploration. Bibliometric Analysis Techniques The bibliometric analysis itself was performed in the Bibliometrix R package with the help of Biblioshiny graphical interface with the assistance of VOSviewer to visualize the network. This combination makes it possible to combine both rigorous quantitative analysis and high-quality visualisation of bibliometric structures. This bibliometric analysis combined performance analysis, science mapping and thematic evolution in a bid to give an overview of the research landscape. The initial indicators that were utilized included the growth of publications, the most prolific sources, and the most influential documents, authors, institutions, and countries, and the measures based on citation were utilized to emphasize the most frequently cited articles and the key sources of publication. Based on this, science mapping methods were used to understand intellectual and social organization of the topic, co-authorship network analysis to identify patterns of collaboration between authors, institutions, and countries, and keyword co-occurrence analysis to identify common themes of research and conceptual connections. Lastly, the thematic mapping and evolution analysis, to categorize the research themes by centrality and density and to distinguish between motor, basic, niche, and emerging themes, was performed to research how scholarly focus has changed between 2023 and 2025. Biblioshiny was chosen because of its transparency, reproducibility, and accessibility, enabling complex bibliometric analyses to be performed without a lot of coding but with maintaining the methodology rigor. These analyses were supplemented with VOSviewer to effortlessly create publication ready network visualizations. Methodological Rigor and Reproducibility In order to increase the reliability, a workflow documenting and replicable data processing and analysis steps were performed. The standardized bibliometric tools, well-designed inclusion criteria, and adapted PRISMA protocol are evident to guarantee that the results can be replicated or further applied by future researchers. The combination of quantitative markers and visual mapping has an overall and methodologically adequate picture of the research landscape. Results and Discussion The section provides and discusses the bibliometric evidence based on the analysis of 282 publications with Scopus indexing in 2023–2025. The findings are arranged in a way that they indicate the performance in the publications, the intellectual organization, social patterns of collaboration and the development of themes. Those findings are incorporated into the discussion of the existing literature on artificial intelligence, and personalized learning in higher education along with theoretical, pedagogical, and institutional implications. Descriptive Characteristics and Publication Trends The most important features of the final dataset are summarized in Table 1 . There were 282 publications that were analyzed and 930 authors made contributions on 57 countries. Mean authors per document is around 3.3 which implies a move towards group study. This interdisciplinary interest in the area of Generative AI and Personalized Learning can be seen by the high number of multi-authored documents. Table 1 Descriptive Characteristics of the Final Dataset Characteristic Value Total publications analyzed 282 Publication period 2023–2025 Total number of authors 930 Number of countries represented 57 Average authors per document 3.3 Dominant authorship pattern Multi-authored publications Research focus Generative AI and Personalized Learning in Higher Education Collaboration trend High level of international and interdisciplinary collaboration Table 2 Metadata Completeness of the Dataset Metadata Field Description Missing Count Missing (%) Status AB Abstract 0 0.00 Excellent C1 Affiliation 0 0.00 Excellent AU Author 0 0.00 Excellent DI DOI 0 0.00 Excellent SO Journal 0 0.00 Excellent DE Keywords 0 0.00 Excellent PY Publication Year 0 0.00 Excellent TI Title 0 0.00 Excellent TC Total Citations 0 0.00 Excellent The number of academic articles published on the topic of generative AI and personalized learning in higher education has grown remarkably between 2023 and 2025. There is an increased interest in research globally, and it can be assumed that it could be accelerated by improvements in AI capabilities (Lachheb et al., 2025 ) and by the normalization of digital pedagogies after the COVID-19 pandemic. The greatest number of publications was in 2024 as Table 1 (Annual Scientific Production) shows, as the educational reforms of the post-pandemic era and the change of policies into promoting innovations in the higher education sector have peaked. The annual scientific production and indicates a steep increase in 2023 to 2025. Though the volume of research output in early 2023 was low, volume of publication was high in 2024 and it kept growing in 2025. This upsurge is correlated with the general increase of using large language models in the educational process and an increase in levels of scholarly interest in the understanding of their pedagogical, ethical, and institutional consequences. The same acceleration patterns are also noted in the general AI-in-education research, especially since important technological breakthroughs and in the global transition to digital learning. Influential Sources and Highly Cited Documents The most relevant sources are discussed in Fig. 3 . The source journal analysis shows that Education and Information Technologies, Computers and Education, and Smart Learning Environments are predominant in the scholarly publication in the given domain. These are Q1 indexed journals that are characterized by the publication of interdisciplinary research that combines AI and educational sciences (Huang et al., 2023 ). These sources are most popular because it is a manifestation of the work of the academic community to prove the effectiveness of AI-based personalized learning tools in education, their ethical aspects, and their technical feasibility. Analysis of the most globally cited documents reveals that the intellectual center of the generative AI and personalized learning studies in tertiary education is conditioned by a few researchers who have the greatest impact in this field. Chan CKY., (2023) overtake all the citations in the world with the fundamental research on the AI-based individual learning experience and adaptive educational systems, which preconditions further research with conceptual and pedagogic support. Complementing this pedagogical focus, Kamalov F., (2023) contribute extensively to ethical, sustainability, and governance aspects of the adoption of AI in higher education, which situates generative AI in the wider context of institutions and society as a whole. Rasul T., (2023) has a strong presence of the empirical perspectives as the author of his work on learning analytics and AI-based feedback has to provide evidence-based insights into the efficiency of personalization made possible by generative technologies. The further contribution to the methodological progress is the work of Abulibdeh A., (2024) whose works regarding the use of AI-based educational modeling and predicting the performance of a learner are extensively referenced due to the rigor of their analysis. Lastly, Firat M., (2023) takes the literature a step further to applications that can be applied in practice by analyzing instructional innovation and implemented learning technologies within a real-life educational context. Together, these internationally mentioned contributions depict a developed but quickly developing research environment that brings together pedagogical theory, personalization that is fuelled by analytics, ethical concerns, and applied execution, emphasizing the multidimensional quality of the generative AI research in higher education. Geographic Distribution and International Collaboration Table 3 Geographic Distribution and International Collaboration Country Year Articles AUSTRALIA 2023 5 AUSTRALIA 2024 21 AUSTRALIA 2025 30 CHINA 2023 10 CHINA 2024 48 CHINA 2025 87 GERMANY 2023 5 GERMANY 2024 12 GERMANY 2025 29 USA 2023 2 USA 2024 22 USA 2025 41 SAUDI ARABIA 2023 0 SAUDI ARABIA 2024 18 SAUDI ARABIA 2025 24 The geography of publications shows that the research production in various regions increased significantly in the period between 2023 and 2025, which implies the worldwide dissemination of the interest towards generative AI and personalized learning in higher education. According to Table 3 , the most significant growth in scholarly productivity is exhibited by China, which currently produces 10 publications per year but is projected to increase to 48 publications in 2024, and 87 publications in 2025. Such a high rate of increase is an indication that a large part of the national economy is invested in artificial intelligence research and its implementation in the educational system. On the same note, the trend of publication output in Australia shows a consistent upward trajectory, as it is set to rise to 5 articles in 2023, 21 in 2024, and 30 in 2025, indicating a continuous interest in AI-enhanced educational innovation research in Australia. The US is also experiencing considerable growth, as the number of publications steadily increases and goes up to 22 in 2024 and 41 in 2025, indicating a rapid growth in scholarly interest after the adoption of large language models in the higher education contexts. European contributions are represented notably by Germany, which demonstrates consistent growth from 5 publications in 2023 to 12 in 2024 and 29 in 2025, reflecting increasing regional engagement with generative AI-driven learning research. In parallel, Saudi Arabia emerges as a key contributor from the Middle East, with no publications recorded in 2023 but a rapid rise to 18 publications in 2024 and 24 in 2025. The trend highlights the increased attention given to digital transformation and AI application in higher education by emerging and developing regions. Taken together, these trends suggest that the studies on generative AI and personalized learning are no longer controlled by the historically leading research economies but are getting influenced more by an assortment of countries. The identified growth in different regions also presupposes the further opportunities of international cooperation, transfer of knowledge and cross-cultural attitude to reshape the further research agenda of AI-based personalized learning in the context of higher education. The trend of the productivity of country-level publication output, as shown in Fig. 5 above indicate that there is a steady increase in the productivity of research in all the considered countries over 2023–2025. China has the strongest growth curve with the number of publications beginning at an average in 2023 then rapidly rising to the peak in 2025. This sharp increase makes China the leading source of research on generative AI and personalized learning during the considered period. United States A United States also has a high and steady growth, as the quantity of publications also increases significantly between 2023 and 2024, and further, as of 2025 the number of publications continues to grow due to the accelerating scholarly interest in large language models becoming mainstream in higher education. The case is similar in Australia and Germany whereby there is a smooth yet steady growth in the publication output over the three years. The trend of Australia means that research is continuously being done and early adoption of generative AI use in education, whereas the consistent increase in Germany means that Europe is increasingly involved in this field of research. Conversely, the Saudi Arabian growth pattern is more delayed but significant with little output in 2023 with a steeper increment between 2024 and 2025. This direction enables noting the active development of research contributions by the Middle East in order to meet the national digital transformation plans. On the whole, the graph highlights a growing research environment on a global level, with the already existing and new research economies making more contributions to the body of scholarly discussion regarding generative AI-driven personalized learning in higher education. Table 4 Country-level participation and collaboration frequency in research on generative AI and personalized learning in higher education (2023–2025) Country Frequency CHINA 87 USA 41 AUSTRALIA 30 GERMANY 29 SAUDI ARABIA 24 UNITED ARAB EMIRATES 23 INDONESIA 20 UK 20 SPAIN 18 JORDAN 17 INDIA 16 MALAYSIA 16 SOUTH KOREA 14 OMAN 12 CANADA 10 MEXICO 10 SOUTH AFRICA 10 THAILAND 10 The geographical aspect of the research publication shows that there is the world interest and involvement in the field of Generative AI and Personalized Learning in Higher Education between 2023 and 2025. The People Republic of China is the most active with 87 total publications according to Table 4 , which shows that the country has been putting a lot of money into the integration of AI technologies into the educational environment. The US is the second with 41 publications, as it has been on top technology and education research. Other significant contributors are Australia (30), Germany (29) and Saudi Arabia (24) which exhibit a multi-regional and diverse interest in exploring AI based personalization in learning. Interestingly, the United Arab Emirates (23) and Malaysia (16) are also ranked highly implying that there is a growing regional focus in the Middle East and Southeast Asia on the need to match education systems with the agenda of digital transformation. This is further supported by the availability of countries such as Jordan (17), Oman (12) and South Africa (10) which are some of the countries which are increasingly becoming global players even amongst the emerging economies. This geographical distribution suggests that not only the global field of generative AI in higher education is possible but also there can be the possibility of cross-country collaboration. The connection between national education policy and publication output or the mapping of regional citation impact in future studies would help understand how geopolitical and socio-economic elements influence academic discussion of AI in education. Author Collaboration and Social Structure of the Field The author collaboration network visualized using Biblioshiny in Fig. 6 reveals several distinct clusters of prolific contributors. For example, Zhang Y., Xu X., and Wang X. belong to a strong Asian based cluster and Li Y., Wang S., and Liu C. to another cluster. The fact that there are bridging authors such as Xu X. implies cross-institutional cooperation and the transfer of knowledge. Chen et al. ( 2022 ) note that this type of network organization is necessary to maintain innovation and prevent the formation of silos in regions regarding rapidly evolving research fields. Regarding the social structure approach, the identified patterns of collaboration indicate the transition to networked scholarship, which is especially significant in highly dynamic fields like generative AI. Collaborative research not only facilitates methodological innovation but also supports the development of shared ethical standards and pedagogical frameworks across institutions and regions. Conceptual Structure and Keyword Co-occurrence Thematic mapping of the keywords generated by VOSviewer in Fig. 7 shows that the area is concentrated around a number of dominant ideas: "ChatGPT," “higher education”, “generative AI”, and “academic integrity”. Other new clusters are adaptive learning, personalized feedback, educational technology, and student engagement. These are indicative of the combination of theoretical and application-based questions implying the multifacetedness of the field. The appearance of keywords connected to ethical AI, evaluation, and student agency is an indicator of the transition to pedagogical implication studies instead of technical feasibility research (Tang and Xiao, 2023 ). The prevalence of keywords related to ethics denotes a change in the academic priorities of moving on discussing technological possibilities and focusing on responsible and pedagogically acceptable practices. This shift indicates that the discipline is ceasing to be experimental and begins to be critical of the role of generative AI in the higher education systems. Thematic Mapping and Evolution of Research Topics To further understand the intellectual organization and the development of research interests, a thematic map was created with the help of Keyword Plus co-occurrence clusters according to centrality (relevance) and density (development) as demonstrated in Fig. 8 . The thematic map categorizes themes in four quadrants, which include Motor Themes, Niche Themes, Emerging or Declining Themes, and Basic Themes (Cobo et al., 2011 ). Motor Themes (High Centrality and High Density) Motor themes represent well-developed and essential topics that structure the field. In this study, “student engagement,” “ai-enhanced learning,” “education,” “large language models,” “programming education,” and “motivation” appear as core drivers of research development in this quadrant. These terms imply the persistence of the need to implement AI and generative models to improve pedagogical performance and student-based approaches in higher education. The emergence of LLMs (Large Language Models) and motivation and student engagement signify the emphasis on how the tools such as ChatGPT may affect intrinsic and extrinsic motivation of learners (Smutny and Schreiberova, 2020 ; Yang and Ogata, 2024 ). Niche Themes (High Density, Low Centrality) This quadrant captures specialized but isolated research areas. Themes such as “augmented reality,” “blended learning,” “feedback,” “instruction,” and “generative artificial intelligence (GAI)” are present, suggesting methodological and technological experimentation in specific domains (e.g., medical or STEM education). Although established, these themes are currently playing a minor role in the general educational discourse, possibly because of the application of the domains to them or the level of technological maturity (Zawacki-Richter et al., 2019 ). Emerging or Declining Themes (Low Centrality, Low Density) This quadrant is particularly insightful in signaling early or waning areas. Here, terms like “ethical AI,” “technology acceptance,” “innovative teaching methods,” “e-learning adoption,” and “generative AI tools” appear. These keywords can indicate early investigation of AI acceptance by the users in the educational field or the conclusion of interest in the initial concepts that are now established. Since the recent ChatGPT and AI ethics discussions are still booming, it is more probable that they are just an emerging theme that will be more fully incorporated over time (Dwivedi et al., 2023 ; Mhlanga, 2023 ). Basic and Transversal Themes (High Centrality, Low Density) Keywords in this quadrant are highly connected yet underdeveloped. This includes “chatgpt,” “higher education,” “artificial intelligence,” “digital literacy,” “student perceptions,” and “bibliometric analysis.” Their position highlights their significance in the research discourse, yet it also presupposes a theoretical synthesis or an increase in empirical studies. To illustrate, even though ChatGPT is currently gaining worldwide popularity, the study of its pedagogical patterns, learning outcomes, and long-term effects is still at an initial stage (Aljanabi and Noorbehbahani, 2023 ). On the same note, UTAUT2, which is used in this case, is not fully utilized as a model to explain AI acceptance in academic literature. Table 5 Major Thematic Clusters in Research on Generative AI and Personalized Learning in Higher Education (2023–2025) Thematic Cluster Quadrant (Thematic Map) Dominant Keywords Research Focus Educational Interpretation Ethics, Integrity, and Responsible AI Emerging or Declining Themes ChatGPT, academic integrity, plagiarism, ethical AI, assessment Ethical risks and governance of generative AI use in higher education Highlights growing concern over trust, authorship, and assessment validity, prompting calls for policy frameworks and ethical guidelines AI-Enhanced Personalized Learning Motor Themes personalized learning, adaptive learning, AI tutor, feedback, learning analytics AI-driven personalization of content, feedback, and learning pathways Reflects the pedagogical shift toward learner-centered and adaptive instructional models supported by generative AI Student Engagement and Motivation Motor Themes student engagement, motivation, self-regulated learning, learning experience Impact of generative AI on learner participation and autonomy Suggests increasing emphasis on how AI tools influence motivation, agency, and active learning in higher education Technology Integration and Digital Pedagogy Niche Themes educational technology, digital learning, blendedlearning, instructional design Integration of generative AI into teaching practices and learning environments Indicates institutional and instructional experimentation with embedding AI into curriculum and digital pedagogy Large Language Models and AI Capabilities Basic Themes generative AI, large language models, natural language processing, conversational AI Technical affordances of LLMs in educational contexts Represents foundational exploration of AI capabilities that underpin pedagogical and personalized learning applications Adoption, Acceptance, and Policy Readiness Emerging or Declining Themes technology acceptance, AI literacy, governance, higher education policy Institutional and user readiness for generative AI adoption Signals emerging interest in organizational, cultural, and policy factors shaping sustainable AI integration To complement the thematic evolution map, Table 5 synthesizes the dominant thematic clusters identified through keyword co-occurrence analysis and interprets their educational significance. As shown in Fig. 8 , research themes are positioned according to their relevance (centrality) and level of development (density), distinguishing between motor, basic, niche, and emerging themes. Motor themes like student engagement and AI-enhanced learning have central and well-developed roles, whereas simple and emerging themes like generative AI, ethical AI, and technology acceptance represent the basis of the research field that is still developing in the research sphere. Conclusion, Implications, and Future Research Conclusion This study provides a comprehensive bibliometric overview of research on generative AI and personalized learning in higher education between 2023 and 2025. The study traces intellectual, social, and conceptual structures of this fast-evolving research field by systematically reviewing 282 publications indexed in Scopus with Bibliometrix (through Biblioshiny) and VOSviewer. The results indicate a steep increase in the academic output, which demonstrates the systemic change in the way of teaching, learning, and assessing that occurs due to the large language models and generative artificial intelligence tools in higher education. These findings show that the study in the area has developed a step further than those initial technical experimentation to a more advanced stage of inquiry which includes the pedagogical, ethical, and institutional inquiry. The overwhelming themes of student engagement, AI-enhanced personalized learning, and academic integrity highlight both the dual nature of generative AI as an enabler of learner-centered education and a source of new issues in the area of trust, governance, and educational values. Moreover, the spread of publications and collaboration networks all over the world confirms the international meaningfulness of generative AI in higher education and the development of various contributions of research on different regions. Overall, this bibliometric review can provide important up-to-date information regarding the role of generative AI in changing personalized learning at a higher educational establishment and may serve as the conceptual framework to support the development of the theory, research, and practice regarding educational technology. Theoretical Implications From a theoretical perspective, this research is relevant to the literature since it clarifies how the field of generative AI studies overlaps with the existing educational theories on personalized learning, student engagement, and self-regulated learning. The prevalence of motivation, engagement, and learner autonomy-related themes imply that generative AI tools are now viewed as more of pedagogical agents than helpful technology. This transition demands theoretical models of the integration of human-AI interaction with theories of learning, that go beyond the traditional models of adaptive, personalized learning. The thematic mapping also indicates the lack of established technology acceptance and learning theories in describing the adoption of generative AI in higher education. Theoretical studies might be able to incorporate more explicitly in the future various frameworks like UTAUT, self-determination theory, and cognitive load theory to elucidate the reactions of learners and instructors to AI-provided personalization. The research fills these theoretical gaps, and therefore it contributes to the creation of theory-based stronger researches on generative AI in education. Pedagogical and Institutional Implications The results have significant implications to the practice of higher education. The increased attention to the development of AI-driven personalized learning and engagement imply that generative AI could facilitate differentiated instruction, formative feedback, and self-directed learning under the provision of the decent pedagogical design. Educators are encouraged to move beyond ad hoc use of AI tools and instead integrate generative AI within curricula through clearly defined learning objectives, assessment strategies, and AI literacy development. The fact that ethics-related and governance-related themes are quite prominent at the institutional level speaks of the necessity of holistic policies concerning the responsible use of AI. Academic integrity, transparency, and data privacy should be clearly laid out by universities, and at the same time, faculty should be assisted by professional development programs. The generative AI adoption needs institutional preparedness at both the level of technological infrastructure and the level of cultural and pedagogical alignment to be sustainable and equitable to implement it. Policy Implications From a policy perspective, it is indicated that the allocation of research output in the world is in such a way that the generative AI in the education sector depends upon the digital transformation policy of the nation and the regulatory conditions. The policymakers must think of implementing loose but sensible structures that would strike equilibrium between innovation and ethical protection. The creation of unified standards of AI usage in higher education, specifically its use in assessment, accreditation, and quality assurance, may be further enhanced with the help of international cooperation and exchange of knowledge. Limitations Despite its contributions, this study has several limitations. First, the analysis will be conducted using only Scopus-indexed publications, which will not cover any important studies listed in other databases like Web of Science or ERIC. Second, bibliometric analysis uses metadata as its basis and does not offer a complete analysis of a literature, preventing the evaluation of the quality of methodology and the empirical rigor of specific studies. Third, the publication period from 2023-2025 restricts the discussion of long-term trends in AI-assisted personalized learning. Future Research Directions Future research can be enhanced in several ways. To begin with, it would be better to expand the analysis to a number of databases and time intervals to get a wider perspective on the development of the field. Second, the bibliometric knowledge should be supported by empirical research investigating learning outcomes, instructional effectiveness, and equity implications of generative AI-enabled personalization. Third, cross-regional and institutional comparisons may enhance the knowledge of the influence of policy, culture, and infrastructure on the adoption of AI in higher education. Lastly, one of the ways in which future work can be conducted is to investigate ethical work and governance models that facilitate responsible innovation and maintain academic integrity and the quality of education (Lin et al., 2025). Declarations AI Use and Ethical Statement This study employed AI-based tools exclusively for bibliometric analysis, data visualization, and language refinement. Generative AI was not used to generate research data, analytical results, or scholarly interpretations. All interpretations, conclusions, and implications were developed and verified by the authors to ensure academic integrity and compliance with ethical research standards. Author Contribution N.A.N is the sole author of this manuscript and was responsible for the conceptualization and design of the study, data collection and curation, bibliometric analysis using the Bibliometrix R package (Biblioshiny) and VOSviewer, interpretation of results, visualization, and manuscript writing and revision. The author has read and approved the final version of the manuscript. Data Availability The data supporting the findings of this study were obtained from the Scopus database, consisting of bibliographic records of publications related to generative AI and personalized learning in higher education published between 2023 and 2025. Due to Scopus licensing restrictions, the raw bibliographic data cannot be publicly shared. However, the search strategy, inclusion criteria, and data analysis procedures are fully described in the manuscript to ensure transparency and reproducibility. Derived data, including aggregated bibliometric indicators and visualization outputs generated using the Bibliometrix R package (Biblioshiny) and VOSviewer, are available from the author upon reasonable request. References Alharthi A, Alghamdi A, Khan M (2024) Digital transformation in Middle Eastern higher education: The role of generative AI. Educ Inform Technol 29(1):12–24 Aljanabi S, Noorbehbahani F (2023) Investigating the adoption of ChatGPT in higher education: A systematic review. J Educational Comput Res 61(6):1283–1302 Ateş SK, Kaleci F, Erdoğan A (2025) Artificial intelligence in education: A bibliometric analysis. Ahmet Keleşoğlu Eğitim Fakültesi Dergisi 7(1):14–36 Chan CKY, Hu W (2023) Students’ voices on generative AI: Perceptions, benefits, and challenges in higher education. Int J Educational Technol High Educ 20:43 Chen Y, Liu D, Zhang H (2022) Mapping co-authorship networks in AI and education: A VOSviewer analysis. Scientometrics 127(4):3011–3028 Cobo MJ, López-Herrera AG, Herrera-Viedma E, Herrera F (2011) Science mapping software tools: Review, analysis, and cooperative study among tools. J Am Soc Inform Sci Technol 62(7):1382–1402 Durak G, Çankaya S, Özdemir D, Can S (2024) Artificial intelligence in education: A bibliometric study on its role in transforming teaching and learning. Int Rev Res Open Distrib Learn 25(3):219–244 Dwivedi YK, Hughes L, Baabdullah AM et al (2023) Artificial intelligence (AI): Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy. Int J Inf Manag 70:102641 Fortuna A, Prasetya F, Samala AD, Rawas S, Criollo-C S, Kaya D, Nabawi RA (2025) Artificial intelligence in personalized learning: A global systematic review of current advancements and shaping future opportunities. Social Sci Humanit Open 12:102114 Haddaway NR, Page MJ, Pritchard CC, McGuinness LA (2022) PRISMA2020: An R package and Shiny app for producing PRISMA 2020-compliant flow diagrams, with interactivity for optimised digital transparency and open synthesis. Campbell Syst Reviews 18:e1230 Holmes W, Bialik M, Fadel C (2022) Artificial intelligence in education: Promise and implications for teaching and learning. Comput Educ 188:104552 Huang RH, Spector JM, Yang J (2023) AI-powered learning: Challenges and directions. Smart Learn Environ 10(1):1–18 Kasneci E, Sessler K, Küchemann S, Bannert M, Dementieva D, Fischer F, Gasser U, Sailer M (2023) ChatGPT for good? On opportunities and challenges of large language models for education. Learn Individual Differences 103:102274 Lachheb A, Leung J, Abramenka-Lachheb V, Sankaranarayanan R (2025) AI in higher education: A bibliometric analysis, synthesis, and a critique of research. Internet High Educ 67:101021 Lin Z, Jurafsky D, Manning C (2025) Next horizons for language models: Challenges and research directions. Trans Association Comput Linguistics 11:205–224 Merino-Campos C (2025) The impact of artificial intelligence on personalized learning in higher education: A systematic review. Trends High Educ 4(2):17 Mhlanga D (2023) Open AI in education: The responsible and ethical use of ChatGPT towards lifelong learning. Educ Inform Technol 28(4):4493–4511 Mustafa MY, Tlili A, Lampropoulos G et al (2024) A systematic review of literature reviews on artificial intelligence in education (AIED): A roadmap to a future research agenda. Smart Learn Environ 11:59 Noordin NAN, Al-Ali MA, A. Y, Alkhamisi MAHO (2024) Students' intention to learn data analytics using learning management systems in UAE higher education institutions: A study on the technology acceptance model. Asian J Univ Educ 20(3):688–702 OECD (2023) AI and the future of skills, Volume 1: Capabilities and assessments. OECD Publishing Page MJ, McKenzie JE, Bossuyt PM, Boutron I et al (2021) The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ 372:n71 Shahzad MF, Xu S, Javed I (2024) ChatGPT awareness, acceptance, and adoption in higher education: The role of trust as a cornerstone. Int J Educational Technol High Educ 21:46 Smutny P, Schreiberova P (2020) Chatbots for learning: A review of educational chatbots for the Facebook Messenger. Comput Educ 151:103862 Susnjak T (2023) ChatGPT: The end of online exam integrity? J Appl Learn Technol 13(1):45–52 Tang Y, Xiao J (2023) Ethical implications of generative artificial intelligence in teaching and assessment. Comput Educ 200:104819 Vieriu AM, Petrea G (2025) The impact of artificial intelligence (AI) on students’ academic development. Educ Sci 15(3):343 Vorobyeva KI, Belous S, Savchenko NV, Smirnova LM, Nikitina SA, Zhdanov SP (2025) Personalized learning through AI: Pedagogical approaches and critical insights. Contemp Educational Technol 17(2):ep574 Xia Q, Weng X, Ouyang F et al (2024) A scoping review on how generative artificial intelligence transforms assessment in higher education. Int J Educational Technol High Educ 21:40 Yang L, Ogata H (2024) Measuring student engagement in AI-enhanced learning environments: A multidimensional approach. Educational Technol Soc 27(1):95–107 Yusuf A, Pervin N, Román-González M (2024) Generative AI and the future of higher education: A threat to academic integrity or reformation? Evidence from multicultural perspectives. Int J Educational Technol High Educ 21:21 Zawacki-Richter O, Marín VI, Bond M, Gouverneur F (2019) Systematic review of research on artificial intelligence applications in higher education: Where are the educators? Int J Educational Technol High Educ 16:39 Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9113918","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":605614731,"identity":"1e3f3984-6ba0-43b9-ae75-caca11a24764","order_by":0,"name":"Nora Azima Noordin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuUlEQVRIiWNgGAWjYJCCAwwMNgkka0lLYGAj0aLDJGgxZz/+8NCNmvN5/PMbGD/8YLCTI6jFsich4XDOsdvFEscYmCV7GJKNCWoxOJBw4HAO2+3EDUCHSTMwMCc2ENRy/mHD4Zx/50BamH8zMNTXE9ZyI5nhcG7bAZAWNmlwOBDW8gyopS85ccaxxDbLHoPjhkQ4LP3x55xvdon9zYcP3/hRUS1P0BYkwAg034AUDaNgFIyCUTAKcAIASqQ9jBGAsTYAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-9756-8561","institution":"Higher Colleges of Technology","correspondingAuthor":true,"prefix":"","firstName":"Nora","middleName":"Azima","lastName":"Noordin","suffix":""}],"badges":[],"createdAt":"2026-03-13 10:56:06","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9113918/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9113918/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104683455,"identity":"1583775a-6b62-48cd-ab7e-9f62902ff076","added_by":"auto","created_at":"2026-03-16 03:34:09","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":120123,"visible":true,"origin":"","legend":"\u003cp\u003ePRISMA protocol\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/6bfdc4d91f9c51d38eb72886.png"},{"id":104683456,"identity":"40612252-ccf8-4a08-abce-acc2871af4ed","added_by":"auto","created_at":"2026-03-16 03:34:09","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":49745,"visible":true,"origin":"","legend":"\u003cp\u003eAnnual scientific production on generative AI and personalized learning in higher education (2023–2025)\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/c6f9b7616584f5a2542584e7.png"},{"id":104782881,"identity":"c03ca137-cd55-498f-b778-2cc55c228f8b","added_by":"auto","created_at":"2026-03-17 07:57:54","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":103208,"visible":true,"origin":"","legend":"\u003cp\u003eTop 10 most relevant sources publishing research on generative AI and personalized learning in higher education (2023–2025)\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/902343a86b6b125495e1ed26.png"},{"id":104782620,"identity":"9763a6b9-39bd-4c24-ab60-773f71a25717","added_by":"auto","created_at":"2026-03-17 07:57:37","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":99253,"visible":true,"origin":"","legend":"\u003cp\u003eMost locally cited documents in generative AI and personalized learning research in higher education (2023–2025)\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/aa53148587784ab92db2d88d.png"},{"id":104683461,"identity":"5fe3f0d4-192b-46f3-a6cc-143282654517","added_by":"auto","created_at":"2026-03-16 03:34:09","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":81577,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal trends in country-level research output on generative AI and personalized learning in higher education (2023-2025)\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/b2dd70f07ab8288bb08dc30b.png"},{"id":104683459,"identity":"8a76118c-2f31-41d9-b226-c063430aeee8","added_by":"auto","created_at":"2026-03-16 03:34:09","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":136079,"visible":true,"origin":"","legend":"\u003cp\u003eAuthor co-authorship network in research on generative AI and personalized learning in higher education (2023–2025)\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/b4fa07662f4824cea413bdfd.png"},{"id":104782888,"identity":"53dda3ea-c5c9-43cb-8285-f19d192c1ada","added_by":"auto","created_at":"2026-03-17 07:57:55","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":520579,"visible":true,"origin":"","legend":"\u003cp\u003eKeyword Co-occurrence Network (results from VOSviewer)\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/71e0f94c21f715effd189644.png"},{"id":104683458,"identity":"883400a3-c369-457c-b71c-169546c75e01","added_by":"auto","created_at":"2026-03-16 03:34:09","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":188118,"visible":true,"origin":"","legend":"\u003cp\u003eThematic evolution map of research on generative AI and personalized learning in higher education based on keyword co-occurrence (2023–2025).\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/48a3802eee803548f9b46433.png"},{"id":106415069,"identity":"e59d4896-5c20-4aa1-9781-1bf956abfe73","added_by":"auto","created_at":"2026-04-08 10:32:42","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2396094,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9113918/v1/34de8cbe-52f3-4c26-b4b5-68d7c6e935ed.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eMapping the Rise of Generative AI in Personalized Learning: A Global Bibliometric Analysis in Higher Education\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe swift advancement progress of Generative Artificial Intelligence (GenAI), particularly large language models (LLM) like ChatGPT, has brought a complete paradigm shift when it comes to the way digital technologies are being conceptualized and used in the context of higher education. In contrast to earlier types of educational artificial intelligence, which were mostly limited to predictive analytics, rule-based systems or adaptive paths, generative AI systems have the ability to generate human like text, descriptions, responses and teaching materials in real-time. This is an ability that has made GenAI potentially an extraordinary power in the teaching, learning, and assessment processes in institutions of higher learning (Dwivedi et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Mhlanga, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003ePersonalized learning has been known to be a very important pedagogical strategy to deal with the diversity among learners to increase the engagement and also help them to make learning self-regulated in higher learning. Personalization has traditionally been supported using learning analytics, intelligent tutoring systems, and adaptive learning platforms, as well as changing the content or pacing relative to the predefined rules or a learner (Zawacki-Richter et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Generative AI is a qualitative extension of this paradigm since these systems are dynamically capable of producing customized explanations, formative feedback, and learning materials that shall be in accordance with the needs of the individual learner. According to recent research, GenAI-led personalization can potentially facilitate learner agency and motivation, as well as transform the design and the role of an educator (Huang et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe interest in generative AI in the academic setting has intensely increased since the open release and popularization of large language models in 2023. A broad variety of uses, such as AI-assisted feedback, automated evaluation, academic writing assistance, tutoring, and learning analytics-based personalization have been discussed by researchers. Meanwhile, this fast spread has brought serious issues regarding the academic integrity, ethically sound application, transparency, and the future implications of generative AI on learning and educational principles (Tang and Xiao, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Aljanabi and Noorbehbahani, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These two narratives, namely, innovation and risk, have informed the current discourse of GenAI in higher education.\u003c/p\u003e \u003cp\u003eDespite this growing body of research, a few review articles and concept papers have discussed artificial intelligence in education more generally, many of them date back to a time when generative AI was at early stages of development, or AI is understood as a uniform group regardless of the type of model (Mustafa et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). As a result, systematic knowledge regarding the development of research at the intersection of generative AI and personalized learning in higher education is limited over the past years, such as the post-LLM emergence period.\u003c/p\u003e \u003cp\u003eThe bibliometric analysis is an objective and strict method to fill this gap by estimating the pattern of publications, the sources of influence, the network of collaborations, and thematic development in a field of research. Unlike narrative or systematic reviews, a bibliometrical technique can be used to determine macro-level trends and scholarly output structures, which are built upon large quantities of scholarly output (Cobo et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). The use of bibliometric methods on the fast growing literature on generative AI and personalized learning is especially useful in capturing general trends in research and inform future investigation.\u003c/p\u003e \u003cp\u003eAccordingly, this study performs a thorough bibliometric search of publications Scopus-indexed between 2023 and 2025 that discuss the topic of generative AI and personalized learning in post-secondary education. The study was examined through the Bibliometrix R package (through Biblioshiny) and VOSviewer which analyzes the growth of publications, leading journals and documents, patterns of collaboration in the world, and development of themes. A modified PRISMA protocol is used to make the data selection and processing transparent and replicable (Page et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis study makes three key contributions. First, it offers a current and targeted mapping of the research at the interface of generative AI and personalized learning in a crucial time of technological transformations in higher education. Second, it determines prevailing and new research topics, showing a change in the initial technical orientation to the pedagogical, ethical, and institutional issues. Third, it provides researchers, educators, and policymakers with evidence-based insights to design, govern, and evaluate higher education generative AI-enabled personalized learning.\u003c/p\u003e \u003cp\u003eThis bibliometric review contributes to the current knowledge on how AI generated (generative AI) is transforming personalized learning and facilitates the creation of theoretically, empirically and ethically responsible and pedagogically significant AI uses in higher education.\u003c/p\u003e"},{"header":"Literature Review","content":"\u003cp\u003eArtificial intelligence has been increasingly embedded in higher education through applications such as learning analytics, intelligent tutoring systems, and adaptive learning platforms. The previous systems were majorly based on predictive models and predetermined instructional rules to facilitate learning. Generative AI, and especially large language models (LLMs) like ChatGPT, is a notable rejection of such methods, as these systems are capable of producing natural language output, explanations, and feedback, which are reminiscent of human interaction (Yusuf et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAccording to the recent study, the spread of generative AI tools in the higher education sector is going viral and is used in academic writing assistance, tutoring, feedback, and assessment creation (Xia et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Researchers highlight that generative AI systems cannot be seen as productivity tools only but can be considered a new category of educational technology that can influence learner cognition, engagement, and instructional practices. Simultaneously, the issues of academic integrity, excessive dependence on AIs, and the loss of critical thinking have been given the center stage (Shahzad et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe literature increasingly conceptualizes generative AI as a pedagogical agent rather than a passive tool. This has led to the demand of new instructional designs, assessment designs, and faculty development plans, that aligns AI utilization with the learning outcomes and educational values (Chan et al., 2023). Consequently, studies on generative AI in higher education have grown at an alarming rate, though conceptually varied and experimental and methodological in nature.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePersonalized Learning and AI-Driven Adaptation\u003c/h2\u003e \u003cp\u003ePersonalized learning is not a new idea that has been promoted as an effective learning strategy to meet the diversity of learners and enhance learning outcomes in higher education. Learning analytics, adaptive learning systems, and intelligent tutoring systems have been used to support traditional methods of personalization, which varies content and pacing depending on learner data (Fortuna et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Such systems are, nevertheless, usually based on decision rules which are established and restricted interaction modalities.\u003c/p\u003e \u003cp\u003eGenerative AI provides a novel aspect of customized learning by offering real-time and conversational, contextual modification. Empirical research proposes that AI-based learning assistants can personalize explanations and create practice resources as well as formative feedback to different needs and preferences of individual learners (Yang and Ogata, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Holmes et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). These capabilities can be heard by constructivist and learner-centered pedagogies that focus on active learning and self-regulation.\u003c/p\u003e \u003cp\u003eHowever, another problem that may arise as a result of AI-based personalization is noticeable in the literature as well. The existence of many issues, including algorithmic bias, unequal quality of personalization, and transparency of AI decision-making, casts their doubts on equity and pedagogical validity (Merino-Campos, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Vieriu and Petrea, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Therefore, researchers put forward that personalization made possible by generative AI should be informed by teaching principles and backed with empirical data about the effectiveness of the learning process as opposed to technological novelty alone (Noordin et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eEthical, Pedagogical, and Institutional Considerations\u003c/h3\u003e\n\u003cp\u003eThe fast implementation of generative AI in higher education has heightened ethical, pedagogical, and institutional controversies. Issues related to ethics that are discussed in the literature are the privacy of data, responsibility, bias in the work of AI, and how the use of AI can affect academic integrity (Vorobyeva et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In assessment contexts, generative AI challenges traditional notions of authorship and originality, prompting reconsideration of assessment design and evaluation criteria.\u003c/p\u003e \u003cp\u003ePedagogically, scholars believe in applying generative AI to curricula in a manner that facilitates AI literacy, critical analysis, and responsible application. Instead of limiting the use of AI tools, most scholars recommend teaching methods that directly involve students in thinking about the advantages and weaknesses of the content created by AI. At the institutional level, universities are coming up with policies and governance structures to control the use of AI in teaching and learning. These policies are region-specific and they capture the difference in regulation environment, technological preparedness and the culture that embraces AI (OECD, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Based on the literature, it is stressed that effective and responsible deployment of generative AI needs institutional incentive, faculty preparation, and consistency between technological framework and instructional objectives.\u003c/p\u003e\n\u003ch3\u003eBibliometric Studies on AI and Education\u003c/h3\u003e\n\u003cp\u003eThe use of bibliometric studies has been important in indexing the research trends of artificial intelligence in education. The present research analyzed the publication growth, collaboration networks, and thematic structures in the fields of learning analytics, intelligent tutoring systems, and educational data mining (Durak et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Ateş et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). These papers are a rich source of macro-level data on the development of the research on AI.\u003c/p\u003e \u003cp\u003eNonetheless, the majority of available bibliometric reviews are not dated after the popularization of generative AI or are simply conceptualizing AI in general and vague terms. Consequently, they fail to record the unique research dynamics related to large language models or their presence in personalized learning conditions. Few recent reviews consider generative AI explicitly and none of them specifically address higher education settings or thematic development in the post-2023 period.\u003c/p\u003e\n\u003ch3\u003eResearch Gap and Study Positioning\u003c/h3\u003e\n\u003cp\u003eAccording to the literature that has been reviewed, three major gaps are apparent. To start with, there is very limited recent bibliometric review that specifically addresses the topic of generative AI and personalized learning in higher education after the advent of large language models. Second, the literature supports minimal research to determine the way research themes have changed following the ethical, pedagogical, and institutional issues related to generative AI. Third, patterns of collaboration and contributions to research in this developing area across the globe are underexplored.\u003c/p\u003e \u003cp\u003eTo address these gaps, the present study applies a narrow bibliometric search of the Scopus-indexed papers published in 2023\u0026ndash;2025. This analysis combines performance analysis, collaboration mapping, key word co-occurrence and thematic evolution analysis to have a holistic view of the intellectual, social and conceptual frameworks that define research on generative AI and personalized learning in higher education. By doing so, it builds on previous bibliometric efforts and provides current information on what needs to be done to further the development of theory, research, and practice in educational technology.\u003c/p\u003e"},{"header":"Methodology","content":"\u003cp\u003eThis study follows a bibliometric research design, which will identify and synthesize the body of academic literature on the topic of generative AI and personalized learning in higher learning. Bibliometric analysis makes it possible to analyze quantitatively the patterns of publications, the intellectual organization, collaboration patterns, and the development of themes in a research sphere. In order to improve the transparency and replicability of the study, the study adheres to an adapted PRISMA 2020 procedure that is progressively used in bibliometric reviews to report data collection, screening, and inclusion processes.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eData Source and Search Strategy\u003c/h2\u003e \u003cp\u003eThe reason to choose Scopus as the main source of data is its extensive coverage of the peer-reviewed journals and conference proceedings in the field of education, technology, and interdisciplinary research. Scopus is a well-known and trustworthy source of bibliometric research and can be easily used with bibliometric analysis software (Bibliometrix and VOSviewer).\u003c/p\u003e \u003cp\u003eThe search strategy was created to include a literature that covers the combination of generative AI, personalized learning, and higher education. The desired comprehensive search string was built on the base of the key words in the sphere of generative AI technologies (e.g., \u0026ldquo;generative AI,\u0026rdquo; \u0026ldquo;large language models,\u0026rdquo; \u0026ldquo;ChatGPT,\u0026rdquo; \u0026ldquo;AI chatbots\u0026rdquo;), personalized learning concepts (e.g., \u0026ldquo;personalized learning,\u0026rdquo; \u0026ldquo;adaptive learning,\u0026rdquo; \u0026ldquo;intelligent tutoring systems\u0026rdquo;), and higher education contexts (e.g., \u0026ldquo;higher education,\u0026rdquo; \u0026ldquo;university,\u0026rdquo; \u0026ldquo;tertiary education\u0026rdquo;). The search was applied to titles, abstracts, and keywords.\u003c/p\u003e \u003cp\u003eIn order to capture the latest trends, after the massive use of large language models, the search was restricted to articles published in 2023\u0026ndash;2025. The last search took place in Scopus in November 2025 and was consistent in all of the searched records.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eInclusion and Exclusion Criteria\u003c/h3\u003e\n\u003cp\u003eIn order to make sure that the dataset remains relevant and of high quality, there were explicit inclusion and exclusion criteria. The criteria of inclusion were peer-reviewed journal articles and conference proceedings, publications in English, and studies that specifically covered the topic of generative AI and personalized or adaptive learning in a higher education context. Publications were excluded if they editorials, short notes, reviews of books, non-academic documents and those published before 2023 were excluded. Upon the use of these criteria and elimination of duplicate records, a final dataset of 282 publications was left to be analyzed.\u003c/p\u003e\n\u003ch3\u003ePRISMA-Based Data Screening and Selection\u003c/h3\u003e\n\u003cp\u003eThe screening and selection process were documented to an adapted PRISMA 2020. First, 299 records were obtained in Scopus according to the set search string and period of time. Redundant records were also detected and eliminated to make 282 distinct publications. Since the bibliometric analysis is a qualitative study, full-text screening was not carried out but rather relevancy was verified by metadata validation, keywords matching and source checking. The PRISMA flow diagram is adapted, which demonstrates every step of the data selection and cleaning process, making it more transparent in the methodology.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eData Cleaning and Pre-Processing\u003c/h2\u003e \u003cp\u003eBefore the analysis, a systematic cleaning and pre-processing stage was applied to the bibliographic data to give it accuracy and consistency. To ensure that the dataset uploaded to Scopus contains all the necessary metadata fields, the dataset was checked to ensure the coverage of the following fields authorship, publication year, source title, affiliations, keywords, and abstract. Redundant records were eliminated and inconsistency in author names, affiliation and special characters were regularized to prevent fragmentation in network analyses.\u003c/p\u003e \u003cp\u003ePreliminary preprocessing was done through the use of spreadsheet software to look at and rectify the clear formatting errors. The filtered data was subsequently inputted into the R studio software to be prepared further with the help of the Bibliometrix package. Bibliographic areas were all standardized to make them compatible with bibliometric indicators and network studies. Such stage of preprocessing made sure that the final dataset can be analysed effectively in terms of robust performance, collaboration mapping and thematic exploration.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eBibliometric Analysis Techniques\u003c/h2\u003e \u003cp\u003eThe bibliometric analysis itself was performed in the Bibliometrix R package with the help of Biblioshiny graphical interface with the assistance of VOSviewer to visualize the network. This combination makes it possible to combine both rigorous quantitative analysis and high-quality visualisation of bibliometric structures.\u003c/p\u003e \u003cp\u003eThis bibliometric analysis combined performance analysis, science mapping and thematic evolution in a bid to give an overview of the research landscape. The initial indicators that were utilized included the growth of publications, the most prolific sources, and the most influential documents, authors, institutions, and countries, and the measures based on citation were utilized to emphasize the most frequently cited articles and the key sources of publication. Based on this, science mapping methods were used to understand intellectual and social organization of the topic, co-authorship network analysis to identify patterns of collaboration between authors, institutions, and countries, and keyword co-occurrence analysis to identify common themes of research and conceptual connections. Lastly, the thematic mapping and evolution analysis, to categorize the research themes by centrality and density and to distinguish between motor, basic, niche, and emerging themes, was performed to research how scholarly focus has changed between 2023 and 2025.\u003c/p\u003e \u003cp\u003eBiblioshiny was chosen because of its transparency, reproducibility, and accessibility, enabling complex bibliometric analyses to be performed without a lot of coding but with maintaining the methodology rigor. These analyses were supplemented with VOSviewer to effortlessly create publication ready network visualizations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eMethodological Rigor and Reproducibility\u003c/h2\u003e \u003cp\u003eIn order to increase the reliability, a workflow documenting and replicable data processing and analysis steps were performed. The standardized bibliometric tools, well-designed inclusion criteria, and adapted PRISMA protocol are evident to guarantee that the results can be replicated or further applied by future researchers. The combination of quantitative markers and visual mapping has an overall and methodologically adequate picture of the research landscape.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Results and Discussion","content":"\u003cp\u003eThe section provides and discusses the bibliometric evidence based on the analysis of 282 publications with Scopus indexing in 2023\u0026ndash;2025. The findings are arranged in a way that they indicate the performance in the publications, the intellectual organization, social patterns of collaboration and the development of themes. Those findings are incorporated into the discussion of the existing literature on artificial intelligence, and personalized learning in higher education along with theoretical, pedagogical, and institutional implications.\u003c/p\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eDescriptive Characteristics and Publication Trends\u003c/h2\u003e \u003cp\u003eThe most important features of the final dataset are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. There were 282 publications that were analyzed and 930 authors made contributions on 57 countries. Mean authors per document is around 3.3 which implies a move towards group study. This interdisciplinary interest in the area of Generative AI and Personalized Learning can be seen by the high number of multi-authored documents.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive Characteristics of the Final Dataset\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal publications analyzed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e282\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePublication period\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2023\u0026ndash;2025\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal number of authors\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e930\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of countries represented\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage authors per document\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDominant authorship pattern\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMulti-authored publications\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResearch focus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGenerative AI and Personalized Learning in Higher Education\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCollaboration trend\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh level of international and interdisciplinary collaboration\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMetadata Completeness of the Dataset\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetadata Field\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMissing Count\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMissing (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStatus\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAbstract\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAffiliation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAuthor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDOI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSO\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eJournal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKeywords\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePY\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePublication Year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTitle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal Citations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExcellent\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe number of academic articles published on the topic of generative AI and personalized learning in higher education has grown remarkably between 2023 and 2025. There is an increased interest in research globally, and it can be assumed that it could be accelerated by improvements in AI capabilities (Lachheb et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and by the normalization of digital pedagogies after the COVID-19 pandemic. The greatest number of publications was in 2024 as Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (Annual Scientific Production) shows, as the educational reforms of the post-pandemic era and the change of policies into promoting innovations in the higher education sector have peaked.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe annual scientific production and indicates a steep increase in 2023 to 2025. Though the volume of research output in early 2023 was low, volume of publication was high in 2024 and it kept growing in 2025. This upsurge is correlated with the general increase of using large language models in the educational process and an increase in levels of scholarly interest in the understanding of their pedagogical, ethical, and institutional consequences. The same acceleration patterns are also noted in the general AI-in-education research, especially since important technological breakthroughs and in the global transition to digital learning.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eInfluential Sources and Highly Cited Documents\u003c/h2\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe most relevant sources are discussed in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. The source journal analysis shows that Education and Information Technologies, Computers and Education, and Smart Learning Environments are predominant in the scholarly publication in the given domain. These are Q1 indexed journals that are characterized by the publication of interdisciplinary research that combines AI and educational sciences (Huang et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These sources are most popular because it is a manifestation of the work of the academic community to prove the effectiveness of AI-based personalized learning tools in education, their ethical aspects, and their technical feasibility.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAnalysis of the most globally cited documents reveals that the intellectual center of the generative AI and personalized learning studies in tertiary education is conditioned by a few researchers who have the greatest impact in this field. Chan CKY., (2023) overtake all the citations in the world with the fundamental research on the AI-based individual learning experience and adaptive educational systems, which preconditions further research with conceptual and pedagogic support.\u003c/p\u003e \u003cp\u003eComplementing this pedagogical focus, Kamalov F., (2023) contribute extensively to ethical, sustainability, and governance aspects of the adoption of AI in higher education, which situates generative AI in the wider context of institutions and society as a whole. Rasul T., (2023) has a strong presence of the empirical perspectives as the author of his work on learning analytics and AI-based feedback has to provide evidence-based insights into the efficiency of personalization made possible by generative technologies. The further contribution to the methodological progress is the work of Abulibdeh A., (2024) whose works regarding the use of AI-based educational modeling and predicting the performance of a learner are extensively referenced due to the rigor of their analysis. Lastly, Firat M., (2023) takes the literature a step further to applications that can be applied in practice by analyzing instructional innovation and implemented learning technologies within a real-life educational context. Together, these internationally mentioned contributions depict a developed but quickly developing research environment that brings together pedagogical theory, personalization that is fuelled by analytics, ethical concerns, and applied execution, emphasizing the multidimensional quality of the generative AI research in higher education.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eGeographic Distribution and International Collaboration\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eGeographic Distribution and International Collaboration\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCountry\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYear\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eArticles\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAUSTRALIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAUSTRALIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAUSTRALIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCHINA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCHINA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e48\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCHINA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e87\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGERMANY\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGERMANY\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGERMANY\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e29\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUSA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUSA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUSA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSAUDI ARABIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSAUDI ARABIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSAUDI ARABIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe geography of publications shows that the research production in various regions increased significantly in the period between 2023 and 2025, which implies the worldwide dissemination of the interest towards generative AI and personalized learning in higher education. According to Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, the most significant growth in scholarly productivity is exhibited by China, which currently produces 10 publications per year but is projected to increase to 48 publications in 2024, and 87 publications in 2025. Such a high rate of increase is an indication that a large part of the national economy is invested in artificial intelligence research and its implementation in the educational system. On the same note, the trend of publication output in Australia shows a consistent upward trajectory, as it is set to rise to 5 articles in 2023, 21 in 2024, and 30 in 2025, indicating a continuous interest in AI-enhanced educational innovation research in Australia. The US is also experiencing considerable growth, as the number of publications steadily increases and goes up to 22 in 2024 and 41 in 2025, indicating a rapid growth in scholarly interest after the adoption of large language models in the higher education contexts.\u003c/p\u003e \u003cp\u003eEuropean contributions are represented notably by Germany, which demonstrates consistent growth from 5 publications in 2023 to 12 in 2024 and 29 in 2025, reflecting increasing regional engagement with generative AI-driven learning research. In parallel, Saudi Arabia emerges as a key contributor from the Middle East, with no publications recorded in 2023 but a rapid rise to 18 publications in 2024 and 24 in 2025. The trend highlights the increased attention given to digital transformation and AI application in higher education by emerging and developing regions.\u003c/p\u003e \u003cp\u003eTaken together, these trends suggest that the studies on generative AI and personalized learning are no longer controlled by the historically leading research economies but are getting influenced more by an assortment of countries. The identified growth in different regions also presupposes the further opportunities of international cooperation, transfer of knowledge and cross-cultural attitude to reshape the further research agenda of AI-based personalized learning in the context of higher education.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe trend of the productivity of country-level publication output, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e above indicate that there is a steady increase in the productivity of research in all the considered countries over 2023\u0026ndash;2025. China has the strongest growth curve with the number of publications beginning at an average in 2023 then rapidly rising to the peak in 2025. This sharp increase makes China the leading source of research on generative AI and personalized learning during the considered period. United States A United States also has a high and steady growth, as the quantity of publications also increases significantly between 2023 and 2024, and further, as of 2025 the number of publications continues to grow due to the accelerating scholarly interest in large language models becoming mainstream in higher education.\u003c/p\u003e \u003cp\u003eThe case is similar in Australia and Germany whereby there is a smooth yet steady growth in the publication output over the three years. The trend of Australia means that research is continuously being done and early adoption of generative AI use in education, whereas the consistent increase in Germany means that Europe is increasingly involved in this field of research. Conversely, the Saudi Arabian growth pattern is more delayed but significant with little output in 2023 with a steeper increment between 2024 and 2025. This direction enables noting the active development of research contributions by the Middle East in order to meet the national digital transformation plans. On the whole, the graph highlights a growing research environment on a global level, with the already existing and new research economies making more contributions to the body of scholarly discussion regarding generative AI-driven personalized learning in higher education.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCountry-level participation and collaboration frequency in research on generative AI and personalized learning in higher education (2023\u0026ndash;2025)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCountry\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFrequency\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCHINA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e87\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUSA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAUSTRALIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGERMANY\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e29\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSAUDI ARABIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUNITED ARAB EMIRATES\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e23\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eINDONESIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUK\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPAIN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJORDAN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e17\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eINDIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMALAYSIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSOUTH KOREA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOMAN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCANADA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMEXICO\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSOUTH AFRICA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHAILAND\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe geographical aspect of the research publication shows that there is the world interest and involvement in the field of Generative AI and Personalized Learning in Higher Education between 2023 and 2025. The People Republic of China is the most active with 87 total publications according to Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, which shows that the country has been putting a lot of money into the integration of AI technologies into the educational environment. The US is the second with 41 publications, as it has been on top technology and education research.\u003c/p\u003e \u003cp\u003eOther significant contributors are Australia (30), Germany (29) and Saudi Arabia (24) which exhibit a multi-regional and diverse interest in exploring AI based personalization in learning. Interestingly, the United Arab Emirates (23) and Malaysia (16) are also ranked highly implying that there is a growing regional focus in the Middle East and Southeast Asia on the need to match education systems with the agenda of digital transformation. This is further supported by the availability of countries such as Jordan (17), Oman (12) and South Africa (10) which are some of the countries which are increasingly becoming global players even amongst the emerging economies.\u003c/p\u003e \u003cp\u003eThis geographical distribution suggests that not only the global field of generative AI in higher education is possible but also there can be the possibility of cross-country collaboration. The connection between national education policy and publication output or the mapping of regional citation impact in future studies would help understand how geopolitical and socio-economic elements influence academic discussion of AI in education.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eAuthor Collaboration and Social Structure of the Field\u003c/h2\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe author collaboration network visualized using Biblioshiny in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e reveals several distinct clusters of prolific contributors. For example, Zhang Y., Xu X., and Wang X. belong to a strong Asian based cluster and Li Y., Wang S., and Liu C. to another cluster. The fact that there are bridging authors such as Xu X. implies cross-institutional cooperation and the transfer of knowledge. Chen et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) note that this type of network organization is necessary to maintain innovation and prevent the formation of silos in regions regarding rapidly evolving research fields.\u003c/p\u003e \u003cp\u003eRegarding the social structure approach, the identified patterns of collaboration indicate the transition to networked scholarship, which is especially significant in highly dynamic fields like generative AI. Collaborative research not only facilitates methodological innovation but also supports the development of shared ethical standards and pedagogical frameworks across institutions and regions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eConceptual Structure and Keyword Co-occurrence\u003c/h2\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThematic mapping of the keywords generated by VOSviewer in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e shows that the area is concentrated around a number of dominant ideas: \"ChatGPT,\" \u0026ldquo;higher education\u0026rdquo;, \u0026ldquo;generative AI\u0026rdquo;, and \u0026ldquo;academic integrity\u0026rdquo;. Other new clusters are adaptive learning, personalized feedback, educational technology, and student engagement. These are indicative of the combination of theoretical and application-based questions implying the multifacetedness of the field. The appearance of keywords connected to ethical AI, evaluation, and student agency is an indicator of the transition to pedagogical implication studies instead of technical feasibility research (Tang and Xiao, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe prevalence of keywords related to ethics denotes a change in the academic priorities of moving on discussing technological possibilities and focusing on responsible and pedagogically acceptable practices. This shift indicates that the discipline is ceasing to be experimental and begins to be critical of the role of generative AI in the higher education systems.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eThematic Mapping and Evolution of Research Topics\u003c/h2\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo further understand the intellectual organization and the development of research interests, a thematic map was created with the help of Keyword Plus co-occurrence clusters according to centrality (relevance) and density (development) as demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e. The thematic map categorizes themes in four quadrants, which include Motor Themes, Niche Themes, Emerging or Declining Themes, and Basic Themes (Cobo et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eMotor Themes (High Centrality and High Density)\u003c/h2\u003e \u003cp\u003eMotor themes represent well-developed and essential topics that structure the field. In this study, \u0026ldquo;student engagement,\u0026rdquo; \u0026ldquo;ai-enhanced learning,\u0026rdquo; \u0026ldquo;education,\u0026rdquo; \u0026ldquo;large language models,\u0026rdquo; \u0026ldquo;programming education,\u0026rdquo; and \u0026ldquo;motivation\u0026rdquo; appear as core drivers of research development in this quadrant. These terms imply the persistence of the need to implement AI and generative models to improve pedagogical performance and student-based approaches in higher education. The emergence of LLMs (Large Language Models) and motivation and student engagement signify the emphasis on how the tools such as ChatGPT may affect intrinsic and extrinsic motivation of learners (Smutny and Schreiberova, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Yang and Ogata, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eNiche Themes (High Density, Low Centrality)\u003c/h2\u003e \u003cp\u003eThis quadrant captures specialized but isolated research areas. Themes such as \u0026ldquo;augmented reality,\u0026rdquo; \u0026ldquo;blended learning,\u0026rdquo; \u0026ldquo;feedback,\u0026rdquo; \u0026ldquo;instruction,\u0026rdquo; and \u0026ldquo;generative artificial intelligence (GAI)\u0026rdquo; are present, suggesting methodological and technological experimentation in specific domains (e.g., medical or STEM education). Although established, these themes are currently playing a minor role in the general educational discourse, possibly because of the application of the domains to them or the level of technological maturity (Zawacki-Richter et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eEmerging or Declining Themes (Low Centrality, Low Density)\u003c/h2\u003e \u003cp\u003eThis quadrant is particularly insightful in signaling early or waning areas. Here, terms like \u0026ldquo;ethical AI,\u0026rdquo; \u0026ldquo;technology acceptance,\u0026rdquo; \u0026ldquo;innovative teaching methods,\u0026rdquo; \u0026ldquo;e-learning adoption,\u0026rdquo; and \u0026ldquo;generative AI tools\u0026rdquo; appear. These keywords can indicate early investigation of AI acceptance by the users in the educational field or the conclusion of interest in the initial concepts that are now established. Since the recent ChatGPT and AI ethics discussions are still booming, it is more probable that they are just an emerging theme that will be more fully incorporated over time (Dwivedi et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Mhlanga, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eBasic and Transversal Themes (High Centrality, Low Density)\u003c/h2\u003e \u003cp\u003eKeywords in this quadrant are highly connected yet underdeveloped. This includes \u0026ldquo;chatgpt,\u0026rdquo; \u0026ldquo;higher education,\u0026rdquo; \u0026ldquo;artificial intelligence,\u0026rdquo; \u0026ldquo;digital literacy,\u0026rdquo; \u0026ldquo;student perceptions,\u0026rdquo; and \u0026ldquo;bibliometric analysis.\u0026rdquo; Their position highlights their significance in the research discourse, yet it also presupposes a theoretical synthesis or an increase in empirical studies. To illustrate, even though ChatGPT is currently gaining worldwide popularity, the study of its pedagogical patterns, learning outcomes, and long-term effects is still at an initial stage (Aljanabi and Noorbehbahani, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). On the same note, UTAUT2, which is used in this case, is not fully utilized as a model to explain AI acceptance in academic literature.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMajor Thematic Clusters in Research on Generative AI and Personalized Learning in Higher Education (2023\u0026ndash;2025)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThematic Cluster\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eQuadrant (Thematic Map)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDominant Keywords\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eResearch Focus\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEducational Interpretation\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEthics, Integrity, and Responsible AI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEmerging or Declining Themes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eChatGPT, academic integrity, plagiarism, ethical AI, assessment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEthical risks and governance of generative AI use in higher education\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHighlights growing concern over trust, authorship, and assessment validity, prompting calls for policy frameworks and ethical guidelines\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAI-Enhanced Personalized Learning\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMotor Themes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003epersonalized learning, adaptive learning, AI tutor, feedback, learning analytics\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAI-driven personalization of content, feedback, and learning pathways\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eReflects the pedagogical shift toward learner-centered and adaptive instructional models supported by generative AI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eStudent Engagement and Motivation\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMotor Themes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003estudent engagement, motivation, self-regulated learning, learning experience\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eImpact of generative AI on learner participation and autonomy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSuggests increasing emphasis on how AI tools influence motivation, agency, and active learning in higher education\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eTechnology Integration and Digital Pedagogy\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNiche Themes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eeducational technology, digital learning, blendedlearning, instructional design\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eIntegration of generative AI into teaching practices and learning environments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eIndicates institutional and instructional experimentation with embedding AI into curriculum and digital pedagogy\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLarge Language Models and AI Capabilities\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBasic Themes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003egenerative AI, large language models, natural language processing, conversational AI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTechnical affordances of LLMs in educational contexts\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRepresents foundational exploration of AI capabilities that underpin pedagogical and personalized learning applications\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAdoption, Acceptance, and Policy Readiness\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEmerging or Declining Themes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003etechnology acceptance, AI literacy, governance, higher education policy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eInstitutional and user readiness for generative AI adoption\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSignals emerging interest in organizational, cultural, and policy factors shaping sustainable AI integration\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTo complement the thematic evolution map, Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e synthesizes the dominant thematic clusters identified through keyword co-occurrence analysis and interprets their educational significance. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, research themes are positioned according to their relevance (centrality) and level of development (density), distinguishing between motor, basic, niche, and emerging themes. Motor themes like student engagement and AI-enhanced learning have central and well-developed roles, whereas simple and emerging themes like generative AI, ethical AI, and technology acceptance represent the basis of the research field that is still developing in the research sphere.\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusion, Implications, and Future Research","content":"\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study provides a comprehensive bibliometric overview of research on generative AI and personalized learning in higher education between 2023 and 2025. The study traces intellectual, social, and conceptual structures of this fast-evolving research field by systematically reviewing 282 publications indexed in Scopus with Bibliometrix (through Biblioshiny) and VOSviewer. The results indicate a steep increase in the academic output, which demonstrates the systemic change in the way of teaching, learning, and assessing that occurs due to the large language models and generative artificial intelligence tools in higher education.\u003c/p\u003e\n\u003cp\u003eThese findings show that the study in the area has developed a step further than those initial technical experimentation to a more advanced stage of inquiry which includes the pedagogical, ethical, and institutional inquiry. The overwhelming themes of student engagement, AI-enhanced personalized learning, and academic integrity highlight both the dual nature of generative AI as an enabler of learner-centered education and a source of new issues in the area of trust, governance, and educational values. Moreover, the spread of publications and collaboration networks all over the world confirms the international meaningfulness of generative AI in higher education and the development of various contributions of research on different regions.\u003c/p\u003e\n\u003cp\u003eOverall, this bibliometric review can provide important up-to-date information regarding the role of generative AI in changing personalized learning at a higher educational establishment and may serve as the conceptual framework to support the development of the theory, research, and practice regarding educational technology.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTheoretical Implications\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFrom a theoretical perspective, this research is relevant to the literature since it clarifies how the field of generative AI studies overlaps with the existing educational theories on personalized learning, student engagement, and self-regulated learning. The prevalence of motivation, engagement, and learner autonomy-related themes imply that generative AI tools are now viewed as more of pedagogical agents than helpful technology. This transition demands theoretical models of the integration of human-AI interaction with theories of learning, that go beyond the traditional models of adaptive, personalized learning.\u003c/p\u003e\n\u003cp\u003eThe thematic mapping also indicates the lack of established technology acceptance and learning theories in describing the adoption of generative AI in higher education. Theoretical studies might be able to incorporate more explicitly in the future various frameworks like UTAUT, self-determination theory, and cognitive load theory to elucidate the reactions of learners and instructors to AI-provided personalization. The research fills these theoretical gaps, and therefore it contributes to the creation of theory-based stronger researches on generative AI in education.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePedagogical and Institutional Implications\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe results have significant implications to the practice of higher education. The increased attention to the development of AI-driven personalized learning and engagement imply that generative AI could facilitate differentiated instruction, formative feedback, and self-directed learning under the provision of the decent pedagogical design. Educators are encouraged to move beyond ad hoc use of AI tools and instead integrate generative AI within curricula through clearly defined learning objectives, assessment strategies, and AI literacy development.\u003c/p\u003e\n\u003cp\u003eThe fact that ethics-related and governance-related themes are quite prominent at the institutional level speaks of the necessity of holistic policies concerning the responsible use of AI. Academic integrity, transparency, and data privacy should be clearly laid out by universities, and at the same time, faculty should be assisted by professional development programs. The generative AI adoption needs institutional preparedness at both the level of technological infrastructure and the level of cultural and pedagogical alignment to be sustainable and equitable to implement it.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePolicy Implications\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFrom a policy perspective, it is indicated that the allocation of research output in the world is in such a way that the generative AI in the education sector depends upon the digital transformation policy of the nation and the regulatory conditions. The policymakers must think of implementing loose but sensible structures that would strike equilibrium between innovation and ethical protection. The creation of unified standards of AI usage in higher education, specifically its use in assessment, accreditation, and quality assurance, may be further enhanced with the help of international cooperation and exchange of knowledge.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLimitations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDespite its contributions, this study has several limitations. First, the analysis will be conducted using only Scopus-indexed publications, which will not cover any important studies listed in other databases like Web of Science or ERIC. Second, bibliometric analysis uses metadata as its basis and does not offer a complete analysis of a literature, preventing the evaluation of the quality of methodology and the empirical rigor of specific studies. Third, the publication period from 2023-2025 restricts the discussion of long-term trends in AI-assisted personalized learning.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFuture Research Directions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFuture research can be enhanced in several ways. To begin with, it would be better to expand the analysis to a number of databases and time intervals to get a wider perspective on the development of the field. Second, the bibliometric knowledge should be supported by empirical research investigating learning outcomes, instructional effectiveness, and equity implications of generative AI-enabled personalization. Third, cross-regional and institutional comparisons may enhance the knowledge of the influence of policy, culture, and infrastructure on the adoption of AI in higher education. Lastly, one of the ways in which future work can be conducted is to investigate ethical work and governance models that facilitate responsible innovation and maintain academic integrity and the quality of education (Lin et al., 2025).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAI Use and Ethical Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study employed AI-based tools exclusively for bibliometric analysis, data visualization, and language refinement. Generative AI was not used to generate research data, analytical results, or scholarly interpretations. All interpretations, conclusions, and implications were developed and verified by the authors to ensure academic integrity and compliance with ethical research standards.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contribution\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eN.A.N is the sole author of this manuscript and was responsible for the conceptualization and design of the study, data collection and curation, bibliometric analysis using the Bibliometrix R package (Biblioshiny) and VOSviewer, interpretation of results, visualization, and manuscript writing and revision. The author has read and approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data supporting the findings of this study were obtained from the Scopus database, consisting of bibliographic records of publications related to generative AI and personalized learning in higher education published between 2023 and 2025. Due to Scopus licensing restrictions, the raw bibliographic data cannot be publicly shared. However, the search strategy, inclusion criteria, and data analysis procedures are fully described in the manuscript to ensure transparency and reproducibility. Derived data, including aggregated bibliometric indicators and visualization outputs generated using the Bibliometrix R package (Biblioshiny) and VOSviewer, are available from the author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAlharthi A, Alghamdi A, Khan M (2024) Digital transformation in Middle Eastern higher education: The role of generative AI. Educ Inform Technol 29(1):12\u0026ndash;24\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAljanabi S, Noorbehbahani F (2023) Investigating the adoption of ChatGPT in higher education: A systematic review. J Educational Comput Res 61(6):1283\u0026ndash;1302\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAteş SK, Kaleci F, Erdoğan A (2025) Artificial intelligence in education: A bibliometric analysis. Ahmet Keleşoğlu Eğitim Fak\u0026uuml;ltesi Dergisi 7(1):14\u0026ndash;36\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChan CKY, Hu W (2023) Students\u0026rsquo; voices on generative AI: Perceptions, benefits, and challenges in higher education. Int J Educational Technol High Educ 20:43\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Y, Liu D, Zhang H (2022) Mapping co-authorship networks in AI and education: A VOSviewer analysis. Scientometrics 127(4):3011\u0026ndash;3028\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCobo MJ, L\u0026oacute;pez-Herrera AG, Herrera-Viedma E, Herrera F (2011) Science mapping software tools: Review, analysis, and cooperative study among tools. J Am Soc Inform Sci Technol 62(7):1382\u0026ndash;1402\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDurak G, \u0026Ccedil;ankaya S, \u0026Ouml;zdemir D, Can S (2024) Artificial intelligence in education: A bibliometric study on its role in transforming teaching and learning. Int Rev Res Open Distrib Learn 25(3):219\u0026ndash;244\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDwivedi YK, Hughes L, Baabdullah AM et al (2023) Artificial intelligence (AI): Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy. Int J Inf Manag 70:102641\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFortuna A, Prasetya F, Samala AD, Rawas S, Criollo-C S, Kaya D, Nabawi RA (2025) Artificial intelligence in personalized learning: A global systematic review of current advancements and shaping future opportunities. Social Sci Humanit Open 12:102114\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaddaway NR, Page MJ, Pritchard CC, McGuinness LA (2022) PRISMA2020: An R package and Shiny app for producing PRISMA 2020-compliant flow diagrams, with interactivity for optimised digital transparency and open synthesis. Campbell Syst Reviews 18:e1230\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolmes W, Bialik M, Fadel C (2022) Artificial intelligence in education: Promise and implications for teaching and learning. Comput Educ 188:104552\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang RH, Spector JM, Yang J (2023) AI-powered learning: Challenges and directions. Smart Learn Environ 10(1):1\u0026ndash;18\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKasneci E, Sessler K, K\u0026uuml;chemann S, Bannert M, Dementieva D, Fischer F, Gasser U, Sailer M (2023) ChatGPT for good? On opportunities and challenges of large language models for education. Learn Individual Differences 103:102274\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLachheb A, Leung J, Abramenka-Lachheb V, Sankaranarayanan R (2025) AI in higher education: A bibliometric analysis, synthesis, and a critique of research. Internet High Educ 67:101021\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin Z, Jurafsky D, Manning C (2025) Next horizons for language models: Challenges and research directions. Trans Association Comput Linguistics 11:205\u0026ndash;224\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMerino-Campos C (2025) The impact of artificial intelligence on personalized learning in higher education: A systematic review. Trends High Educ 4(2):17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMhlanga D (2023) Open AI in education: The responsible and ethical use of ChatGPT towards lifelong learning. Educ Inform Technol 28(4):4493\u0026ndash;4511\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMustafa MY, Tlili A, Lampropoulos G et al (2024) A systematic review of literature reviews on artificial intelligence in education (AIED): A roadmap to a future research agenda. Smart Learn Environ 11:59\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoordin NAN, Al-Ali MA, A. Y, Alkhamisi MAHO (2024) Students' intention to learn data analytics using learning management systems in UAE higher education institutions: A study on the technology acceptance model. Asian J Univ Educ 20(3):688\u0026ndash;702\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOECD (2023) AI and the future of skills, Volume 1: Capabilities and assessments. OECD Publishing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePage MJ, McKenzie JE, Bossuyt PM, Boutron I et al (2021) The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ 372:n71\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShahzad MF, Xu S, Javed I (2024) ChatGPT awareness, acceptance, and adoption in higher education: The role of trust as a cornerstone. Int J Educational Technol High Educ 21:46\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmutny P, Schreiberova P (2020) Chatbots for learning: A review of educational chatbots for the Facebook Messenger. Comput Educ 151:103862\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSusnjak T (2023) ChatGPT: The end of online exam integrity? J Appl Learn Technol 13(1):45\u0026ndash;52\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang Y, Xiao J (2023) Ethical implications of generative artificial intelligence in teaching and assessment. Comput Educ 200:104819\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVieriu AM, Petrea G (2025) The impact of artificial intelligence (AI) on students\u0026rsquo; academic development. Educ Sci 15(3):343\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVorobyeva KI, Belous S, Savchenko NV, Smirnova LM, Nikitina SA, Zhdanov SP (2025) Personalized learning through AI: Pedagogical approaches and critical insights. Contemp Educational Technol 17(2):ep574\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXia Q, Weng X, Ouyang F et al (2024) A scoping review on how generative artificial intelligence transforms assessment in higher education. Int J Educational Technol High Educ 21:40\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang L, Ogata H (2024) Measuring student engagement in AI-enhanced learning environments: A multidimensional approach. Educational Technol Soc 27(1):95\u0026ndash;107\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYusuf A, Pervin N, Rom\u0026aacute;n-Gonz\u0026aacute;lez M (2024) Generative AI and the future of higher education: A threat to academic integrity or reformation? Evidence from multicultural perspectives. Int J Educational Technol High Educ 21:21\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZawacki-Richter O, Mar\u0026iacute;n VI, Bond M, Gouverneur F (2019) Systematic review of research on artificial intelligence applications in higher education: Where are the educators? Int J Educational Technol High Educ 16:39\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Higher Colleges of Technology","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Generative artificial intelligence, Personalized learning, Higher education, Bibliometric analysis, Large language models","lastPublishedDoi":"10.21203/rs.3.rs-9113918/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9113918/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe rapid development of Generative Artificial Intelligence (GenAI), especially large language models, has heightened academic attention on its role in incentivizing personalized learning in higher education. The current paper will include a thematic bibliometric survey of the Scopus-indexed articles published by 2023 and 2025 to trace the changing research environment at the intersection of generative AI and personalized learning. Through the Bibliometrix R package through Biblioshiny and VOSviewer, the analysis is performed to analyze the development of publications, journals and documents with the greatest impact, patterns of international collaboration, and the development of research topics. The results demonstrate a dramatic rise in the number of research and a great involvement of the world, and also the change in the scholarly perspective of research towards pedagogic, ethical and institutional issues. Such themes of research are AI-facilitated personalized learning, student engagement, academic honesty, and responsible usage of AI. This study can inform researchers, educators, and policymakers in the timely manner by synthesizing the global research trends, so that they can design, govern, and assess generative AI-enhanced personalized learning in higher education.\u003c/p\u003e","manuscriptTitle":"Mapping the Rise of Generative AI in Personalized Learning: A Global Bibliometric Analysis in Higher Education","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-16 03:34:04","doi":"10.21203/rs.3.rs-9113918/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"41bb1344-dd37-4d38-a71e-20ea17f4e3af","owner":[],"postedDate":"March 16th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":64452679,"name":"Artificial Intelligence and Machine Learning"}],"tags":[],"updatedAt":"2026-03-16T03:34:04+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-16 03:34:04","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9113918","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9113918","identity":"rs-9113918","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}