ChatGPT as a Digital Tool in the Transformation of Digital Teaching Competence: A Systematic Review

Abstract

In recent years, the use of tools based on artificial intelligence, such as ChatGPT, has begun to play a relevant role in education, particularly in the development of teachers’ digital competence. However, its impact and the implications of its integration in the educational environment still need to be rigorously analysed. This study aims to examine the role of ChatGPT as a digital tool in the transformation and strengthening of teachers’ digital competence, identifying its advantages and limitations in pedagogical practices. To this end, a systematic literature review was carried out in four academic databases: Web of Science, Scopus, ERIC and Google Scholar. Eighteen relevant articles addressing the relationship between the use of ChatGPT and professional teacher development were selected. Among the main findings, it was identified that this technology can contribute to the continuous updating of teachers, facilitate the understanding of complex content, optimise teaching planning, and reduce the burden of repetitive tasks. However, challenges related to technology dependency, the need for specific training, and the ethics of its educational application were also noted. The results of this study suggest that the use of ChatGPT in education should be approached from a critical and informed perspective, considering both its benefits and limitations. Empirical studies are recommended to evaluate its real impact in different educational contexts and the implementation of teacher training strategies that favour its responsible and effective use in the classroom.

1. Introduction

Digital transformation in the field of education has made it essential to constantly update the Digital Teaching Competence, understood as the set of skills, knowledge, and attitudes that allow teachers to effectively integrate technologies into the teaching–learning process [1]. This competence is not limited to the technical handling of tools but also encompasses the ability to use them in a pedagogical and ethical manner, adapted to the needs of the educational context. It also involves the development of critical and creative skills to solve problems, design innovative learning environments, and respond to constant technological changes. In this sense, several studies have highlighted the urgency of training teachers to meet the challenges of digital education [2,3]. It is also stressed that this competence must evolve towards a more reflective and critical dimension, oriented not only towards the functional use of technology, but also towards its application for educational purposes [4]. This implies that teachers need to assess the impact of digital resources on learning, make informed decisions about their use, and promote inclusive, collaborative and student-centred practices.

The European Commission proposes the European Framework for Digital Competence (DC) in Education (DigCompEdu), which sets out levels and key areas such as information literacy, digital communication, and digital content creation, all of which are essential for teaching in modern environments [5]. Furthermore, this concept is closely linked to critical digital literacy, which involves the ability to understand, analyse, and critically evaluate digital content and the environments in which digital technologies are produced and consumed [6].

Recent studies agree that initial teacher education often has significant gaps in teachers’ technological preparation, which affects their confidence and effectiveness in using ICTs [7]. Research shows that many trainee teachers have a basic level of DC, which reinforces the need for more robust training programmes.

The development of DTC should start in initial teacher education (ITE), as this facilitates the effective integration of ICT into future professional practice [8]. However, recent studies show that student teachers, at the end of their training, only reach a basic level in the assessed dimensions of ICT [7]. On the other hand, research carried out in the university context has shown differences in the level of DC according to variables such as gender, teaching experience and familiarity with ICT, which highlights the need for personalised training programmes [9].

Teacher training plays a key role in the acquisition of this competence, ensuring that educators not only master the use of technology, but can also apply it effectively and reflectively in their pedagogical practices [10]. A strong background in DC, combined with a critical approach to emerging technologies, is indispensable to ensure the responsible and effective integration of tools such as artificial intelligence [AI] in the classroom [11]. This involves not only incorporating advanced technologies, but also rethinking pedagogical approaches to enhance collaborative, personalised and inclusive learning, integrating tools and strategies from the outset. The integration of these technologies should not only be seen as an evolution of traditional methods, but as a paradigm shift that redefines the way that teaching–learning processes are conceived and developed. From this perspective, the concept of computational thinking is introduced to refer to the ability to understand the functioning of AI tools and their applicability in education [12]. The relationship of this notion to our research lies in the evaluation of teacher readiness to integrate these technologies in teaching, highlighting both their potential and their limitations. AI technologies are among the most widely used tools in education. In education, AI tools such as ChatGPT are making an exponential impact, standing out for their potential for both students and learners. In this sense, ChatGPT is an artificial intelligence technology tool because it combines advanced algorithms, efficiency, and a practical approach to problem solving and task assistance, all through natural language-based interaction. AI is defined as a field of science and engineering that deals with the computational understanding of what is commonly referred to as intelligent behaviour, as well as the creation of artefacts that exhibit such behaviour [13]. That is, we can understand AI as the ability of a machine to perform tasks that, if performed by a human, would require intelligence, such as learning, problem solving, and decision making. AI technology tools have been widely recognised as a key tool for improving DC in the educational context, particularly in teacher training and professional development. This technology, based on artificial intelligence, not only facilitates access to information and resources, but also enables teachers to perform essential tasks more efficiently. According to Elsayary [14], these tools can generate content, assist in lesson planning, and provide real-time feedback, which considerably optimises teachers’ work.

ChatGPT has established itself as a key element in the transformation of teachers’ digital competence (TDC), due to its ability to optimise teaching processes, personalise learning experiences, and foster student autonomy [15]. The ChatGPT tool, with its advanced natural language processing capabilities, has positioned itself as an innovative tool in education. Its potential lies not only in the generation of coherent and contextualised responses, but also in its ability to help teachers in key areas of their professional practice. For example, for the creation of digital content, it can help design teaching materials for teachers to apply in their classrooms. In this regard, it is worth noting that ChatGPT learns to generate text from a large corpus of pre-existing data, without a conscious or adaptive understanding of the content discussed. This technology, which operates by recognising language patterns and structures, is designed to produce coherent and contextualised responses to the instructions received. In the area of assessment, it facilitates the automatic generation of questionnaires, rubrics, and feedback tailored to the individual needs of learners, thus promoting more inclusive teaching. Likewise, in the context of teaching, understood as the process of learning facilitation and pedagogical mediation, ChatGPT acts as a virtual assistant, allowing teachers to concentrate on tasks of greater pedagogical value, such as personalised attention, the design of innovative strategies, and the analysis of individual needs in the classroom. Far from replacing the teacher, ChatGPT functions as a catalyst that amplifies the teacher’s ability to address the challenges of contemporary education, freeing up time and resources to strengthen the human interactions essential for deep learning [16].

The educational potential of ChatGPT has been highlighted in recent studies as a tool to support both instructional design and the improvement of DC [17]. In this sense, ChatGPT allows for the creation of tailored educational content, alternative explanations, rubrics, and customised teaching materials, improving competence in resource creation [18]; ChatGPT facilitates lesson planning, the diversification of methodologies (such as project-based learning or flipped classroom), and the personalisation of learning through differentiated activity suggestions, promoting student-centred teaching and its ability to generate instant explanations and examples, making it an ally for teachers seeking to adapt to diverse learning paces and styles [19]. However, it can also generate questionnaires, rubrics, and automatic feedback, supporting more agile and personalised formative assessment. These functionalities enable teachers to monitor their students’ progress more efficiently, provide timely feedback, and allow students to interact with AI technologies in research, writing, and problem-solving tasks. This use promotes critical thinking, self-regulation, and the development of advanced digital literacy, aligning with the competence to empower learners [15].

The impact of tools such as ChatGPT on the teaching–learning process is significant. Their implementation facilitates content personalisation, automated feedback generation, and support for students outside school hours through virtual tutorials, promoting autonomous and continuous learning. These capabilities help teachers optimise time spent on administrative tasks, allowing them to focus on more complex and creative pedagogical activities [14,20,21]. Moreover, their use contributes to the development of new digital and pedagogical competences in teachers, transforming traditional teaching into a more dynamic and interactive experience [16,22,23].

However, it is essential to bear in mind that the inappropriate use of these technological tools can negatively affect the development of critical thinking and problem-solving skills. In addition, ethical concerns related to plagiarism and academic integrity arise due to the ease with which automated texts are generated [24,25]. These risks underline the need for clear policies and effective detection tools, as well as the need to provide teachers with comprehensive digital skills training to enable them to use artificial intelligence critically and responsibly [20,26].

ChatGPT represents a key tool for transforming TDC, optimising educational processes, personalising learning experiences, and fostering learner autonomy [19]. However, its effective implementation requires balancing technological innovation with pedagogical criteria, ensuring that artificial intelligence complements, but does not replace, the critical and reflective work of the educator. This is the only way to guarantee an ethical and sustainable use of these technologies, favouring meaningful learning and the development of DC within a framework of educational responsibility [27].

Therefore, the purpose of this study is to conduct a systematic review of the literature to explore the impact of ChatGPT in the educational field, specifically in the development of TDC. This methodology allows for a rigorous and objective analysis of the available evidence, providing a comprehensive view of the use of this tool in various educational contexts [28]. The systematic review is justified by its ability to synthesise previous research findings, identify emerging patterns and gaps in the knowledge, and provide informed recommendations for practice and future research. This approach has been widely validated in educational and technological studies, establishing itself as a robust methodological strategy for assessing the impact of technological innovations on teaching and learning. The interest in ChatGPT lies in its growing adoption in education, driven by its free availability, ease of use, and ability to generate immediate, coherent, and contextualised responses. These attributes make it a potentially transformative tool for pedagogical practices. Several authors have pointed out its applications, benefits, challenges, and limitations in education [29,30], which reinforce the relevance of a systematic review that gathers and critically analyses this scattered information.

While there are different versions of ChatGPT (such as GPT-3.5 and GPT-4), this study focuses on the basic functions common to all of them, such as coherent text generation, academic writing support, automated feedback, and dialogue simulation. Therefore, the findings obtained are generalisable and replicable, even with future or equivalent versions of the model, ensuring the validity and currency of the conclusions. Thus, the choice of a systematic review is supported by its suitability to critically analyse the impact of emerging technologies such as ChatGPT on teacher education, and by its potential to guide the development of educational policies, training strategies, and innovative pedagogical practices in the digital age.

2. Objectives

In this sense, the aim of this study is to analyse the impact of the use of ChatGPT in the transformation of TDC, evaluating its benefits, challenges, and levels of application according to disciplines and educational stages, in order to propose strategies for its effective integration in the processes of teacher training and practice.

In order to achieve this objective, the following research questions were formulated to guide the research:

RQ1: What is the current state of scientific research on the use of ChatGPT as a tool for developing TDC in education?

RQ2: What are the main uses of ChatGPT identified by teachers in their professional development and educational activities in the classroom?

RQ3: What recommendations have been proposed to integrate ChatGPT effectively into teaching practices and classroom dynamics?

3. Methodology

The present research was conducted through a systematic review of the literature, following the guidelines established by the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) Statement, which provides a recognised and rigorous methodological framework for the planning, execution, and reporting of systematic reviews [31]. This methodological choice ensures the transparency, replicability, and quality of the review process, which are fundamental elements to consolidate the credibility of the study.

The systematic review is particularly suitable for this work, as it provides a comprehensive, reliable, and up-to-date overview of the impact of ChatGPT on the development of digital teaching competence. By synthesising and critically analysing the available empirical evidence, this methodology contributes to both the advancement of academic knowledge and to the identification of practical solutions to current challenges in education [32].

To strengthen the analysis of the data collected, advanced technological tools were employed. First, VOSviewer software (Version 1.6.20) was used for the analysis of co-occurrence networks and the visualisation of thematic relationships between publications. In addition, the statistical software R 4.0 was incorporated through the Biblioshiny application, part of the Bibliometrix package, which allowed a detailed bibliometric analysis to be carried out and facilitated the interpretation of the results in terms of trends, key authors, relevant sources, and thematic evolution. This tool was widely used in previous research to perform bibliometric and network analyses due to its ability to clearly and intuitively visualise maps of co-occurrences, citations, and relationships between the authors or institutions. Numerous studies have validated its usefulness in the field of education, particularly in bibliometric studies related to artificial intelligence, by facilitating the identification of emerging trends. In this regard, some investigations have explored the evolution of AI research in education, highlighting VOSviewer’s value for exploratory analysis and the identification of new research directions [33]. Others have used it to map research in the university context, revealing the key relationships between AI, educational quality, and digital competences in higher education [34].

Overall, the adoption of a systematic methodology supported by consolidated theoretical frameworks and the use of analysis tools recognised in the scientific literature ensure the validity, reliability, and relevance of the findings. This sound methodological basis allows us to offer informed recommendations for the pedagogical integration of ChatGPT and its role in strengthening TDC.

3.1. Search Strategy

This review examines the impact of ChatGPT on digital teaching competence. Accordingly, a set of keywords was used to systematically search for relevant literature in four databases covering various fields such as education, social sciences, and computer science: Web of Science, Scopus, ERIC, and Google Scholar. The selection of these databases for this systematic review was made considering their relevance and thematic coverage in relation to the objective of the study. Web of Science and Scopus were selected as high-quality, multidisciplinary databases that index peer-reviewed research in various disciplines, including education and technology. These databases are recognised for their comprehensiveness and quality criteria, ensuring that the articles included meet rigorous academic standards. The ERIC (Education Resources Information Center) was included specifically for its focus on education, making it an essential source for identifying studies related to teaching competencies. Google Scholar was used as a complement to identify grey literature and relevant studies that may not be indexed in the aforementioned databases, ensuring broader coverage and reducing publication bias. This strategic combination of databases ensures a systematic, comprehensive, and relevant search to address the objective of the study. In September 2024, a comprehensive search was conducted using the following search string: ((‘ChatGPT’ OR ‘Artificial Intelligence’) AND (‘digital competence’ OR ‘teacher education’) AND (‘teachers’)). The selection of general concepts, such as ‘artificial intelligence’, ‘digital competence’, and ‘teacher education’, are chosen because they are widely recognised and necessary to establish a clear theoretical framework. These concepts allow the study to be linked to previous research and ensure its relevance in the academic field. However, care has been taken to contextualise them specifically to the field of education and teacher education, to avoid the analysis being too broad or generic. Both Spanish and English were applied in the title, abstract, and keyword fields of the databases, in order to refine the results obtained. In order to ensure the timeliness and relevance of the findings, the inclusion criterion was set as a selection criterion for scientific articles published between 2022 and 2024. This temporal delimitation is justified by the fact that ChatGPT was launched to the public in November 2022, marking the beginning of its effective integration in various educational contexts. Therefore, the studies published in this period reflect experiences, research, and analysis directly related to the actual use of this tool in contemporary educational settings. Limiting the search to this time range allows us to more accurately capture the recent and specific impact of ChatGPT on the development of DC in teaching, avoiding the inclusion of studies prior to the existence of this technology or research that does not consider its current characteristics. The initial search identified 438 scientific publications.

3.2. Inclusion and Exclusion Criteria

To ensure the selection of relevant articles, eligibility criteria were established. These criteria help to narrow the scope of the search and ensure that the articles included are relevant and effectively contribute to the advancement of our knowledge in the research area.

To be eligible for inclusion in the systematic review, studies had to meet the following criteria:

• Scientific articles published between 2022 and 2024.
• Articles written in English or Spanish.
• Papers in peer-reviewed scientific article format.
• Research specifically focused on the topic of study.
• Descriptive, predictive, and correlation studies.

Studies were excluded if they met any of the following criteria:

• Scientific articles outside the specified range of years (2022–2024).
• Articles published in languages other than English or Spanish.
• Other studies (conference proceedings, book chapters, dissertations...).
• Research that does not focus on the specific field of study.
• Papers that are systematic reviews of the literature.

3.3. Assessment of the Methodological Quality of the Scientific Articles

The methodological quality of the selected studies was assessed using the Johanna Briggs Checklist (JBI), a tool designed for the critical analysis of scientific studies. This checklist allows the validity of studies to be assessed on the basis of a number of criteria covering key aspects of methodology and reported results [35]. Each study was independently reviewed by two reviewers, ensuring an objective analysis and minimising the potential bias in the assessment. To determine whether a study met the minimum quality standards, it was essential to meet at least four of the six criteria listed (see

Table 1. Checklist for critical analysis of studies.

Items	Yes	No
Clarity in the formulation of the research problem: Is the formulation of the research problem clearly defined in the selected studies on the use of ChatGPT to transform teachers’ digital competence?	✓	X
Adequate study design to address the research questions: Is the study design adequate to address the research questions related to transforming teachers’ digital competence through ChatGPT?	✓	X
Reliability of data collection methods: Are the data collection methods used in the studies serve to evaluate the effectiveness of ChatGPT in teacher education reliable?	✓	X
Adequate data analysis: Does the data analysis conducted in the studies determine the impact of ChatGPT on teachers’ digital competence appropriate?	✓	X
Implementation of strategies to reduce bias: Do studies implement effective strategies to reduce bias in assessing the impact of ChatGPT on teacher transformation?	✓	X
Relevance and consistency of findings: Are the findings of the studies relevant and consistent with the objectives and results presented on the influence of ChatGPT on teachers’ digital competence?	✓	X

). These criteria are given in Table 1 below.

Of the studies assessed, two were excluded because they did not meet the minimum requirement of four criteria, which compromised the credibility of their findings. The main weaknesses observed in these studies were related to the lack of adequate design and the lack of strategies to reduce potential bias.

The JBI Checklist was selected because of its systematic approach to assessing the internal validity and methodological reliability of prevalence studies, which is one of the main objectives of this research. Its use ensures that only studies with sufficient methodological rigour are considered for the final analyses and conclusions. The use of the JBI tool is directly aligned with the principles of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework, which guides this systematic review. PRISMA sets standards for the transparent assessment and presentation of reviews, and the integration of the JBI tool reinforces these principles by providing a standardised mechanism for assessing the quality of included studies. In this context, JBI contributes to the construction of a robust and reliable analysis, ensuring that the conclusions derived from the review are valid and consistent with the stated objectives.

3.4. Selection of the Scientific Articles

The initial search identified 438 articles in the databases. After eliminating duplicates (n = 123), the titles and abstracts of the remaining 315 articles were reviewed to assess their relevance according to the inclusion and exclusion criteria, and 205 articles were eliminated. The remaining 110 articles were assessed by an exhaustive review by the authors and according to their methodological quality, with 90 studies eliminated by the authors as they did not fully fit the field of research and a further 2 due to deficiencies in the robustness of their methodologies. Finally, 18 articles were included in the systematic review. The process of study selection is illustrated in the flow chart in Figure 1.

The selected articles can be viewed in

Table 2. Articles included in the systematic review.

Authors	Country	Type of Study	Participants	Results/Implications	Limitations
Iqbal et al. [36]	Malaysia	Qualitative	University teaching staff	Most do not use it in the classroom, due to lack of training, but they are aware of its benefits. Teachers need to improve their training.	The study had a small sample size of only 20 participants, which restricts the generalisability of the results; it did not explore students’ perceptions of ChatGPT use in higher education, nor did it assess how this technology is being used in university teaching and learning.
Nguyen [37]	Vietnam	Mixed	University teaching staff	They are excited about its implementation in the classroom, but emphasise the need for more professional training. Benefits in creating learning resources and helps reduce their workload.	The sample was limited to 20 English teachers at a single university, which restricts the generalisability of the results; furthermore, although the participants were already using ChatGPT, they were still in the process of familiarising themselves with the tool, which limited the depth of their responses.
Ali et al. [38]	Saudi Arabia	Quantitative	University teaching staff	ChatGPT can help to improve students’ academic skills and encourage motivation and interest in the classroom. This requires adequate teacher training.	A limitation of the study was the absence of qualitative data that would have enriched the results. In addition, further quantitative and qualitative exploration of how ChatGPT influences new learning styles and strategies is needed.
Allehyani and Algamd [39]	Saudi Arabia	Quantitative	Early childhood teachers	There is a gap in TDC to apply AI. ChatGPT offers numerous pedagogical possibilities and enriches teaching practice.	The main limitation of the study is its exclusively quantitative approach, based on a self-administered questionnaire, which restricts the depth and richness of the findings on teachers’ perceptions. This methodology prevents us from capturing the complexity of implementing ChatGPT in teaching English as a second language.
Urazbayeva et al. [40]	Kazakhstan	Quantitative	University teaching staff	ChatGPT improves teachers’ AI literacy. It also offers benefits such as creative lesson planning, automated assessments and personalised learning experiences.	The sample was small and geographically limited, which restricts the generalisability of the results. In addition, the intervention was of short duration and did not include a control group, which precludes attributing the observed changes exclusively to the intervention. It also highlights the limitation of using ChatGPT due to its outdated database, which could lead to misinformation, especially in areas such as science.
Firat [41]	Turkey	Qualitative	University teaching staff	The integration of AI in education offers numerous opportunities to improve learning experiences, personalise teaching, reduce teacher workload and transform the role of educators. However, this shift poses challenges in terms of assessment, digital literacy, and ethical considerations.	The research was based on a small sample of 21 academics and Ph.D. students, and used a single open-ended question for data collection, which restricts the depth and diversity of responses. In addition, all participants were assumed to have a good understanding of English, which may not be representative of wider contexts.
Kaplan-Rakowski et al. [42]	USA	Quantitative	Primary education teaching staff	Teachers express positive perspectives towards the integration of AI into their teaching practice. The more they used it, the more positive their perspectives became. Teachers believe that AI can enhance their professional development and can be a valuable tool for students.	Although the study includes a variety of teachers from different contexts, the sample size is relatively small, which limits the generalisability of the results to the whole population of teachers.
Al-Mughairi and Bhaskar [43]	Oman	Qualitative	University teaching staff	Teachers use ChatGPT to explore innovative educational technologies, personalise their teaching practice, save time, and for professional development. However, they are concerned about reliability, reduced human interaction, privacy, lack of institutional support, and over-reliance on ChatGPT.	The sample was limited to 36 teachers from the 10 UTAS sites in Oman, which restricts the generalisability of the findings to other cultural or educational contexts. Furthermore, by employing a qualitative approach based on interpretative phenomenological analysis (IPA), an in-depth but not necessarily representative understanding of all possible experiences was obtained.
Gladstone [44]	India	Theoretical	Does not specify level of education	It is a valuable tool for teachers that allows them to create content tailored to the needs of students, as well as being used as a virtual assistant that facilitates autonomous learning and personalised tutoring.	Although the study considers the diversity of the sample from different contexts, the small sample size limits the possibility of extrapolating the results to the whole population.
Mondal et al. [45]	India	Theoretical	Does not specify level of education	ChatGPT can serve as a personalised tutor that helps students improve their understanding of the topics covered in class. On the other hand, it has the potential to ease the workload of teachers, helping them focus on more complex aspects of the educational process.	The study was based on the free version of the model, which may have restricted some of its advanced capabilities and more precise functions. On the other hand, the examples used in the study were based on the authors’ personal experiences, without empirical or controlled validation in real educational contexts.
Kayali et al. [46]	Turkey	Quantitative	Higher education	This tool is mainly used as a teaching and learning assistant to improve the teaching and learning process of students.	A convenience sampling method was used, with 84 students. ChatGPT had difficulties understanding ambiguous questions, answering long questions correctly or handling unclear conversational situations.
Rahim et al. [47]	Malaysia	Quantitative	Higher education	The use of AI in classrooms is used for language learning, especially improving their writing skills and facilitating greater motivation to learn the language they are studying.	The sample was small and limited to a single university in Malaysia, which restricts the generalisability of the results; convenience sampling was used, which may introduce bias; only student perceptions were analysed without measuring actual improvements in language skills.
Bitzenbauer [48]	Germany	Quantitative	Secondary education	One study found that it has been used to foster critical thinking skills, especially in subjects such as physics, where students can interact with the chatbot to solve problems and clarify complex concepts.	The study consisted of only two class sessions, which is insufficient to assess sustainable impacts on the development of students’ critical thinking. The study focused on analysing students’ opinions about ChatGPT, without directly measuring skills such as critical thinking or conceptual learning.
Estrellado and Millar [49]	Philippines	Qualitative	Secondary education	This AI provides pedagogical support, helping teachers to personalise teaching and tailor assessments to the individual needs of students.	Exclusive use of an interpretive approach based on textual analysis, reliance on personal experiences that hinder comparisons and generalisations, and a small sample size that limits the applicability and replicability of the findings. However, it is noted that these limitations do not affect the validity of the study, as rigorous standards of qualitative research were followed.
Murgia et al. [50]	Italy	Quantitative	Primary education teaching staff	Teachers use ChatGPT with their students to better understand educational content through responses tailored to their literacy level.	A limitation of the study was the small size of the children’s sample and the use of a paper and a pencil to complete the task. It is proposed that digital devices will be used in future research, as there are differences in readability between printed and digital texts in primary education.
El Sayary [14]	United Arab Emirates	Mixed	Secondary education	Teachers’ perceptions are focused on lesson planning, teaching and learning, and to a lesser extent on assessment and feedback. This results in greater engagement on the part of students.	The sample was 40 participants from public schools in Dubai, preventing a larger sample size.
Wardat [51]	United Arab Emirates	Qualitative	Secondary education	In mathematics education, ChatGPT has been valued as a tool that can improve educational success by providing detailed instructions and assistance in complex topics such as geometry.	The sample was small and focused on early adopters, which limits the generalisability of the findings. In addition, the exclusively qualitative approach lacks quantitative support to objectively measure the impact of ChatGPT. Errors in the mathematical reasoning of the model were also identified, especially in topics such as geometry and limits, which is evidence of a lack of conceptual understanding.
Nguyen [52]	Vietnam	Quantitative	University teaching staff	The participating teachers had little knowledge of ChatGPT. The results indicate the need for adequate training or guidance in the use of ChatGPT, and to be able to implement ChatGPT in teaching and assessment.	The study was based on a limited sample of 43 university teachers, which restricts the generalisability of the results. Furthermore, although open-ended questions were used, the data collected lacked depth in terms of teachers’ specific beliefs and experiences. Finally, it points to the need for future mixed-methods research to provide a more complete picture of the use of this tool in educational contexts.

.

The articles on the use of ChatGPT in education included in the systematic review highlight a geographical and methodological diversity. While most of the studies focus on the pedagogical impact, practical applications and challenges perceived by teachers, there is a common trend towards the need for adequate training and effective integration of this tool in specific educational contexts. For example, studies in Malaysia [38], Vietnam [39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54] and Saudi Arabia [38], Allehyani and Algamd [39] highlight that teachers recognise the benefits of ChatGPT, but face barriers related to lack of training and institutional support. In a more practical approach, research in Kazakhstan [40], Turkey [41], and the USA [41] highlights the potential of ChatGPT for lesson planning, automated assessment, and the personalisation of learning, although challenges in terms of digital literacy and ethical considerations are also identified.

Similarly, studies in Oman [43], India [44,45], and Italy [50] emphasise the use of ChatGPT for developing educational resources and enhancing personalised teaching, although concerns remain about reliability, privacy, and the possible reduction in human interaction in the classroom. In the context of secondary and primary education, research in Germany [48], the Philippines [49], and the United Arab Emirates [14], Wardat [51] highlights its role in fostering critical thinking, language teaching, and personalising learning to the individual needs of learners. However, the variability in methodological approaches, the levels of education analysed, and the implications identified highlight the importance of considering contextual factors, such as geographical and cultural settings, when interpreting the results and planning future research on the integration of ChatGPT in education.

4. Results

The results obtained from the systematic review of the literature are as follows:

First, looking at the temporal distribution of the selected articles (Figure 2), it can be seen that most of the articles were published during 2023, with a total of 15 publications, followed by 2024, with 2 articles, and finally, 2022, with only 1 publication. This pattern reflects the recent emergence of ChatGPT, which was publicly released in late 2022. Thus, the inclusion of 18 articles in this systematic review is justified because it is an emerging field in which a solid base of consolidated literature is not yet available. Nevertheless, this number is representative of the scientific output in the short period of time since the initial dissemination of ChatGPT, allowing us to identify initial trends and key areas of interest. We recognise that, given the nascent nature of this phenomenon, the findings reflect a preliminary picture that is likely to be enriched by future research.

Secondly, looking at the geographical distribution of the selected articles (Figure 3), it can be seen that there is a multitude of countries that have carried out research on this subject, mainly Asian countries such as Saudi Arabia, the United Arab Emirates, India, and Turkey, among many others. On the other hand, other publications from European and American countries such as Germany, Italy, and the United States have been highlighted to a lesser extent. The geographical origin of the articles influences the results by showing the differences in access to technology, the research priorities of the countries, or the cultural or economic factors that may affect the implementation of AI or neural networks.

The analysis of the methodology used in the reviewed studies allows us to understand the current panorama of research on the use of ChatGPT in the classroom, highlighting its strengths and possible areas for improvement. According to the data presented in Figure 3, 50% of the selected studies adopt a quantitative approach, which facilitates obtaining measurable and generalisable results. On the other hand, 27.8% use qualitative methods, providing a deeper and more contextualised understanding of the implications of the use of this technology. In addition, 11.1% of the studies combine both methodologies, which enriches the analysis by integrating numerical and descriptive perspectives. Finally, another 11.1% focus on theoretical studies, laying the groundwork for future proposals and practical implementations of ChatGPT in the classroom. This methodological balance evidences the need to continue diversifying approaches to comprehensively address the challenges and opportunities presented by artificial intelligence in education (Figure 4).

The Sankey diagram, also known as a three-field diagram, is a visualisation technique that makes it possible to show in a clear and understandable way the relationships between the key words of the studies, the countries where the research has been carried out, and the academic journals that have published these works. These connections help to identify patterns and trends in the scientific literature that would not be so evident through a textual or tabular analysis [53]. In particular, the diagram represented in Figure 5 examines the relationship between the central keywords in the published studies related to the use of ChatGPT, the country where the research is concentrated and the academic journal within the scope of the scientific literature. The size of the blocks and the connections in the diagram reflect the relative relevance of each element. For example, the topic of “Generative AI” seems to be highly relevant in the research, which could indicate that these countries are engaged in research on how the application of AI, especially ChatGPT, can generate new and meaningful content on its own, often from minimal parameters provided by a human user. This research is published in journals such as the Journal of Applied Learning & Teaching which focuses on innovative teaching and learning practices. Thus, it can be seen how the articles collected on the use of ChatGPT for the training of DC in teaching have been published both in the country and in journals (Figure 5).

On the other hand, regarding the uses that teachers have for ChatGPT in the classroom, it should be noted that the detailed analysis of the collected articles reveals a wide variety of applications identified after an exhaustive review of the literature (Figure 6). In this regard, the research highlights that teachers mainly apply this AI to professional development, closely linked to teachers’ continuous growth on educational trends, as suggested by recent research on new strategies [54]. Secondly, it is highlighted that it is applied to improve knowledge understanding. This suggests that ChatGPT can be an effective source for clarifying complex concepts, offering detailed explanations, and providing additional educational resources that support both students and teachers in understanding difficult topics. Thirdly, we find the reduction in workload and as a virtual assistant for tutoring. In this regard, these results reflect that ChatGPT can handle repetitive tasks such as basic assessment, provide automated feedback, and answer frequently asked questions from students, freeing up time for teachers to focus on other more important aspects of the educational process [55]. Finally, other important uses such as the development of critical thinking for students, the creation of educational resources for their classrooms, or even pedagogical support are highlighted [56].

Keyword co-occurrence analysis is a fundamental tool in systematic reviews, as it allows for the identification of the main thematic areas and conceptual relationships in a field of research. This type of analysis facilitates the understanding of how key terms cluster around specific themes, revealing recurrent research patterns, and highlighting predominant trends in the literature. In this way, Figure 7 represents a network graph or co-occurrence map of keywords generated with VOSviewer. Each node represents a keyword and the relationship between the terms. The different colours suggest groupings or clusters of related concepts. These clusters represent thematic areas within the overall theme. The thematic areas in this line of research are given as follows:

• Red cluster: Evaluation and processes in the Digital Teacher Transformation.

The assessment of TDC, represented by the red cluster, is closely linked to teachers’ perceptions of ChatGPT adoption, analysed in the blue cluster. This is because teachers’ perceptions of the tool’s impact influence how they adopt and use ChatGPT in their pedagogical practices, which in turn, affects their performance evaluation processes. For example, a teacher who perceives ChatGPT as a beneficial tool is likely to adopt more innovative strategies, and thus be more open to dynamic assessments of their DC.

2.

Blue cluster: Perceptions and impact of ChatGPT integration in education.

The perceptions and impacts analysed in the blue cluster serve as a basis for developing the practical recommendations proposed in the green cluster. Studies on teachers’ experience with ChatGPT and its impact in the classroom generate useful information for designing practical integration strategies. For example, identifying the barriers perceived by teachers in the blue cluster allows the formulation of recommendations that effectively address these challenges.

3.

Green cluster: Practical recommendations for the use of artificial intelligence.

This area is more focused on recommendations on how artificial intelligence, such as ChatGPT, can support teachers in the field of research and in their teaching work. That is, it highlights how AI, in this case, ChatGPT, can be effectively integrated in these contexts through practical recommendations based on scientific research.

Together, the three clusters highlight how ChatGPT adoption can be effectively integrated into teacher digital transformation. The red cluster addresses the metrics and processes needed to assess impact, the blue cluster identifies the perceptions that shape this integration, and the green cluster translates these findings into practical actions. These elements provide a holistic perspective, allowing the article to articulate a clear and structured framework for understanding the role of ChatGPT in education today.

5. Discussions

In line with the discussions presented in this article, the organisation was based on the research questions formulated in the objectives.

RQ1: What is the current state of scientific research on the use of ChatGPT as a tool for developing TDC competence in education?

The use of ChatGPT is becoming widespread in all educational stages, as it has been gaining popularity since its appearance. Therefore, the systematic review of the literature revealed that most of the articles included in this systematic review were published during the year 2023. Since its debut in November 2022, ChatGPT became popular extraordinarily quickly, attracting millions of users in just a few weeks. This rapid rise sparked academic interest in investigating its applications in a variety of fields, resulting in 2023 in the publication of numerous research papers on its use [57]. On the other hand, many Asian countries have been pioneers in adopting advanced technologies to improve education. This AI has been seen as a promising tool to personalise teaching, improve accessibility, and facilitate autonomous learning. These advantages have prompted studies exploring how ChatGPT can improve student productivity and performance [58].

Likewise, it has been observed that most studies have opted for qualitative approaches, as opposed to quantitative, mixed, or even theoretical research, in relation to teachers’ use of ChatGPT. This type of study makes it easier for researchers to adopt a deductive, empirical, and positivist approach to data collection and analysis, in addition to achieving a more representative sample of the area of study [59,60].

RQ2: What are the main uses of ChatGPT identified by teachers in their professional development and educational activities in the classroom?

In response to the second research question, the analysed articles show a diversity of uses of ChatGPT in the classroom, characterised by its flexibility to adapt to different pedagogical purposes. In this sense, several authors have shown that this technological tool not only enriches teaching but also contributes to the professional development of teachers by keeping them updated on the latest research and educational methodologies [61]. This is especially valuable for those seeking to improve their pedagogical skills on an ongoing basis.

A crucial aspect of this technology is its ability to provide detailed and contextualised explanations, which enhances the understanding of complex topics for both teachers and students. Kaplan-Rakowski et al. [42] and Lo [62] highlight that ChatGPT plays a significant role in tasks such as personalised tutoring, guiding students according to their specific needs and reinforcing concepts in various disciplines. These studies provide empirical evidence validating its usefulness in varied educational contexts.

In addition, ChatGPT has been found to optimise lesson planning by generating customised teaching content and activities. Rahman and Watanabe [58] and Urazbayeva et al. [41] report that the use of this technology [39,41] allows teachers to spend less time on administrative and repetitive tasks, such as creating materials, to focus on the more strategic aspects of teaching. This assertion is supported by qualitative and quantitative studies that document increases in teacher efficiency and improvements in teacher job satisfaction.

Regarding the reduction in the teaching workload, ChatGPT has been shown to alleviate monotonous and repetitive tasks, such as correcting exercises or generating questionnaires [37,41]. The development of critical thinking is indispensable in today’s society, where the ability to solve everyday problems is indispensable for students. The creation of educational content also has a strong impact on the main uses of this tool. It is mainly aimed at creating personalised learning materials in an efficient way, tailored to the needs of the learners. This is especially useful in subjects that require the constant adaptation and updating of content [52,63]. Finally, it has also been highlighted that the use of ChatGPT as a virtual assistant for student tutoring and pedagogical support has a great impact on teachers. Its effectiveness may depend on the teacher’s ability to correctly guide its use and take advantage of its functionalities within a wider teaching environment. The authors suggest that teachers need to be adequately trained to effectively integrate AI into their classes and avoid misuse by students [44,63].

RQ3: What recommendations have been proposed to effectively integrate ChatGPT into teaching practices and classroom dynamics?

The clusters identified in the co-occurrence map have made it possible to highlight the fundamental aspects of the use of ChatGPT in education, covering the evaluation of TDC, the perceptions and impacts derived from its integration into pedagogical practices, and practical recommendations aimed at its effective implementation in various educational contexts. In relation to research-based practical recommendations on how to integrate ChatGPT into teaching and educational research, these include the following:

• Improve teacher training in DC through ChatGPT: It is crucial to establish ongoing training programmes that strengthen TDC, with a focus on the use of tools such as ChatGPT to optimise lesson design and classroom interaction. These programmes should focus on the practical use of ChatGPT as a digital assistant in lesson planning, assessment and feedback, thus promoting greater efficiency and personalisation of learning. Training should include the development of skills to manage human–machine collaboration effectively and ethically [16].
• Developing new pedagogical models that integrate ChatGPT as a supplementary tool: It is recommended that pedagogical models be designed that include ChatGPT as a supplementary support, rather than a replacement for traditional methodologies [64]. These models should promote personalised teaching and active learning, where AI facilitates access to resources, the analysis of educational data, and support for individual student needs. Integration must be balanced with direct instruction to ensure that critical skills such as analytical thinking and problem solving are maintained and reinforced [41]. The use of AI is also transforming the educational landscape by offering immersive, personalised, and tailored learning experiences. This approach can be integrated with AI to further personalise teaching and learning, as AI can analyse interaction data and adjust content according to the individual needs of each learner, optimising the educational process [65].
• Implement ethical and regulatory oversight: As ChatGPT becomes more integrated into educational processes, regulatory and ethical frameworks need to be put in place to ensure responsible use. This includes clear policies on data privacy, academic integrity, and the safety in the use of AI tools, ensuring that teachers are trained in the safe and ethical use of these technologies in the classroom [14].
• Investigate the integration of ChatGPT in the development of TDC: It is essential that future research explores how the incorporation of ChatGPT into teaching practice can improve TDC and their adaptation to digital educational environments. Studies should focus on optimising the relationship between teachers and AI technologies, assessing their impact on the quality of the educational process and on the development of skills such as critical digital literacy [6]. These studies should also investigate how AI can diversify assessment methods and improve learning outcomes [41]. On the other hand, the environmental impact is very important to investigate. In this regard, Nishant et al. [66] argued that AI has the potential to reduce energy and resource intensity by facilitating more effective environmental governance, which is particularly relevant in the integration of sustainable digital tools within education. In this context, Aggarwal [67] proposes that combining sustainable development initiatives with AI tools in education can foster a more efficient approach, achieving reduced resource use and promoting a positive environmental impact.

These recommendations aim to ensure the effective and responsible implementation of ChatGPT in the development of TDC, enhancing both their training and educational quality.

6. Conclusions

The study on ChatGPT and its impact on the transformation of TDC reveals a growing landscape in the use of this AI in education. Since its emergence in 2022, it has attracted the interest of researchers from different locations, mainly from Asian countries. Studies highlight that ChatGPT has transformed teaching by offering significant benefits, such as teacher professional development, reduced administrative burden, the personalisation of learning, and improved content planning and creation. Teachers use AI not only to keep up to date with educational trends, but also to deepen their understanding of complex concepts, plan lessons, and provide personalised tutorials. In turn, the review identifies that ChatGPT acts as a virtual assistant, helping teachers to reduce monotony in repetitive tasks such as marking assignments and providing feedback, which optimises the time available to focus on higher-value pedagogical activities. However, the use of ChatGPT raises challenges, including technology dependency, which can compromise the development of critical thinking in both teachers and students. Concerns related to ethics and academic integrity are also highlighted, as ChatGPT can generate texts that, without proper control, could facilitate plagiarism. Given these challenges, this study emphasises the need for adequate training for teachers to use AI critically and ethically, avoiding over-reliance. The recommendations of the systematic review underline the importance of establishing continuous training programmes that strengthen TDC, promoting a practical use of ChatGPT as an assistant in lesson planning, assessment, and personalisation of learning. It is also proposed to design new pedagogical models that integrate ChatGPT in a balanced way, promoting active learning and ensuring the development of fundamental skills such as critical thinking and problem solving. It is also essential to implement regulatory and ethical frameworks that ensure the responsible use of this technology, addressing issues such as data privacy and academic integrity.

It is therefore essential for teachers to acquire adequate training in specific DC, enabling them to use artificial intelligence critically and effectively. This becomes a priority, as it represents one of the greatest challenges for education in the 21st century. The research suggests that AI should be complemented with pedagogical judgement to ensure a positive impact on the educational process and on the development of sustainable and ethical DC. In terms of future recommendations, the study calls for further research into how ChatGPT can enhance TDC and optimise the relationship between teachers and AI technologies.

6.1. Limitations

Despite the significant findings of this study, several limitations are identified that need to be considered. First, the systematic review is based only on studies published between 2022 and 2024 because the launch of this technology tool (ChatGPT) was in 2022. It is critical to collect more long-term data to comprehensively assess the impact and implications of the continued use of ChatGPT in education. This would identify the trends, sustainable benefits, and potential challenges associated with its prolonged implementation. In addition, a long-term analysis would help to understand how this tool influences teacher adaptation, DC development, and student learning outcomes in diverse educational and cultural contexts. Another limitation is that the inclusion of articles published in languages other than English or Spanish could have broadened the perspective and covered underrepresented regions. Most of the selected studies come from specific geographical contexts, mainly Asian countries, which may limit the generalisability of the results to other regions with different educational and cultural contexts. In this sense, it could be expanded by exploring how differences between countries affect the results and uses of AI or neural networks, analysing cultural, technological, or economic barriers to AI implementation in different contexts.

6.2. Implications for Theory and Practice

This study significantly contributes to the theoretical development and educational practice around the integration of AI—in particular ChatGPT—in teacher education and practice. On a theoretical level, the findings broaden and enrich the concept of TDC, proposing a vision that incorporates not only technical skills, but also critical, ethical, and pedagogical dimensions necessary for the reflective use of AI-based technologies. This approach complements existing frameworks such as DigCompEdu by proposing a conceptual update that includes computational understanding and critical digital literacy as key elements in 21st century teacher education.

From a practical perspective, the results have direct implications for teachers, policy makers, and educational institutions. For teachers, the study provides evidence of the value of ChatGPT as a tool to support multiple professional tasks, such as instructional planning, personalisation of learning, formative feedback, and continuous professional development. These applications optimise educators’ time, strengthen their role as mediators of learning, and foster more inclusive and innovative pedagogical practices. For policy makers, this work highlights the urgent need to design continuing teacher education strategies that integrate the use of artificial intelligence from an ethical, critical, and contextualised approach. Such strategies should ensure that teachers not only use these tools but also understand their implications and can exercise informed professional judgement about their integration in different educational settings. Finally, for educational institutions, the findings suggest the desirability of rethinking pedagogical and organisational models to facilitate the effective adoption of emerging technologies. This includes promoting flexible and collaborative learning environments, strengthening digital infrastructure, and establishing regulatory frameworks that ensure the responsible and sustainable use of artificial intelligence in education. In doing so, this study not only contributes to the academic debate on DC and the use of AI in educational contexts, but also offers a practical roadmap to guide their effective, ethical, and pedagogically meaningful implementation.

6.3. Implications for Theory and Practice

From the findings of this study, multiple opportunities are identified to advance theoretical and empirical understanding of the use of generative artificial intelligence, such as ChatGPT 4.5, in the development of TDC. Although recent literature has begun to document its initial impact, there remains a pressing need to generate more robust, contextualised, and longitudinal evidence to guide informed pedagogical and policy decisions.

First, there is an identified need for longitudinal studies to assess the sustained impact of ChatGPT and other AI tools on the development of TDC over time. This approach would help to determine not only the immediate benefits, but also the long-term effects on educational practice and professional development processes. It is also suggested to explore educational contexts that are less represented in the current literature, such as rural schools, early childhood education levels, developing countries, or regions with less technological infrastructure. Investigating these realities would contribute to a more inclusive and contextualised understanding of AI integration in education. Future research could also address in greater depth the socio-technical and pedagogical aspects of human–AI collaboration, especially with regard to didactic decision making, teacher autonomy, and the co-creation of educational content. This approach would allow a better understanding of the limits and possibilities of AI as a pedagogical tool, as well as the factors that modulate its critical appropriation by teachers.

Another emerging focus is the development and validation of specific ethical and regulatory frameworks for the use of AI in education. Issues related to data privacy, the equity of access, algorithmic transparency, and academic integrity require a systematic approach from a multidisciplinary perspective. This also involves assessing teachers’ ethical competences and their preparedness to deal with dilemmas related to the use of automated technologies in the classroom. Finally, there is a need to explore the environmental and sustainability implications associated with the use of AI tools in education. Given the increasing use of computationally intensive systems, it is relevant to investigate how to design environmentally responsible implementation strategies aligned with the sustainable development goals (SDGs).

These lines of research will not only contribute to strengthening the theoretical framework of DC in the age of artificial intelligence, but will also provide a solid empirical basis for guiding educational policies, redesigning training models and ensuring a critical, ethical and transformative integration of these emerging technologies.