Generative AI vs. Teachers: insights from a
literature review
IA generativa versus profesores:
reflexiones desde una revisión de la literatura
Dr. Andres Chiappe. Profesor
Titular. Universidad de la Sabana. Colombia
Dña. Carolina
Sanmiguel. Profesora. Fundación Universitaria
Navarra. Colombia
Dra. Fabiola Mabel
Sáez Delgado. Profesora. Universidad Católica de la
Santísima Concepción. Chile
Received: 2024/07/08 Reviewed 2024/05/28 Accepted: :2024/12/06 Online First: 2024/12/14 Published: 2025/01/07
Cómo citar
este artículo:
Chiappe,
A., Sanmiguel, C., & Sáez Delgado, F. M. (2025). IA generativa versus
profesores: reflexiones desde una revisión de la literatura [Generative AI vs. Teachers: insights from a literature review]. Pixel-Bit. Revista De
Medios Y Educación, 72, 119–137. https://doi.org/10.12795/pixelbit.107046
ABSTRACT
The growing integration of
artificial intelligence in universities is reshaping higher education,
particularly through the use of chatbots and generative language models. This article conducts a
literature review, applying PRISMA guidelines to 155 peer-reviewed articles, to
examine the advantages, limitations, and pedagogical applications of AI
compared to human teaching. Three main scenarios of impact on educational
practices were identified: a) Loss of certain traditional aspects of teaching,
such as exclusive information transmission and reporting tasks, b)
Transformation of roles, including control over educational content and the
didactic contract, c) Emergence of new elements, such as personalized learning
and innovative evaluation approaches. Despite its potential to automate
processes and save time, chatbots cannot replicate essential human qualities
like empathy and adaptability. Therefore, their optimal integration requires
thorough pedagogical analysis to balance innovation with educational
effectiveness. This work is valuable for researchers, educators, and
instructional designers seeking to understand how to leverage AI without
compromising teaching quality. It represents a crucial step toward the
development of AI integration strategies grounded in solid pedagogical principles.
RESUMEN
La creciente integración
educativa de la inteligencia artificial está reconfigurando la educación
superior, especialmente a través del uso de chatbots y
modelos de lenguaje generativo. Este artículo realiza una revisión de la
literatura, aplicando las directrices PRISMA a 155 artículos revisados por
pares, para examinar las ventajas, limitaciones y aplicaciones pedagógicas de
la IA en comparación con la enseñanza humana. Se identificaron tres principales
escenarios de impacto en las prácticas educativas: a) Pérdida de ciertos
aspectos tradicionales de la enseñanza, como la transmisión exclusiva de
información y tareas de reporte, b) Transformación de roles, incluyendo el
control sobre contenidos educativos y el contrato didáctico, c) Emergencia de
nuevos elementos, como la personalización del aprendizaje y enfoques
innovadores en la evaluación. A pesar de
su potencial para automatizar procesos y ahorrar tiempo, los chatbots no replican cualidades humanas esenciales como la
empatía y la adaptabilidad. Por ello, su integración óptima requiere análisis
pedagógicos profundos que equilibren innovación y efectividad educativa. Este
trabajo es valioso para investigadores, docentes y diseñadores educativos
interesados en entender cómo aprovechar la IA sin comprometer la calidad de la
enseñanza. Representa un paso crucial hacia estrategias de incorporación de IA
basadas en principios pedagógicos sólidos.
KEYWORDS· PALABRAS CLAVES
Generative Artificial Intelligence; Teacher Practices;
Educational Innovation; Higher Education; Pedagogical Transformation; Chatbot
Applications in Education.
Inteligencia Artificial Generativa; Prácticas
Docentes; Innovación Educativa; Educación Superior; Transformación Pedagógica;
Aplicaciones de Chatbots en Educación.
1. Introduction
In recent times it has become more and more common or frequent
to hear about pilot implementation experiences of chatbots in education, as
part of a growing and increasingly complex trend of incorporating digital
technologies to support teaching and learning (Chen et al., 2023;
Tlili et al., 2023).
In this regard, Salvagno et al. (2023), mention that
chatbots are programs capable of generating a specific conversation with
people, through natural language processing. Chatbots, which can link text as
well as voice, can recognize expressions, understand perspectives, and offer
insights from ongoing feeding or training processes based on their users'
responses and interactions. In other words, chatbots are considered a software
tool that allows interaction with users regarding a certain topic or also on a
specific domain in a natural and conversational way through text and voice (Smutny &
Schreiberova, 2020). They have been used
for many different purposes, in a wide range of domains, and education has not
been the exception.
The few and most recent investigations in this regard
and the information available in the press and academic networks indicate that
there is a lot of confusion and fear regarding the use of these digital tools,
mainly related to plagiarism (King, 2023) and, in general, to
the loss of relevance of many of the learning and evaluation activities that
have traditionally been provided to students (Surahman & Wang,
2022).
In this scenario, it is vitally important to offer a
reflective approach from a pedagogical perspective on this matter, so that it
is useful for researchers and educators, and thus identify its possibilities
and main risks for its proper implementation in the framework of higher
education. The path to understanding, at least in an incipient way, the
potential and risks of using chatbots in education, it seems that almost
everything is still to be done, according to what is indicated in Figure 1,
where the research panorama is shown in this subject, published in
peer-reviewed journals indexed in Scopus.
Figure 1
Published articles on
“chatbots AND education” in peer-reviewed journals indexed in Scopus
Figure 1 highlights that investigative interest in the
use of chatbots in education has grown exponentially over the past 5 years.
However, the number of articles published per year is still relatively low,
with an average of just over 100 articles per year in the last four years.
These findings suggest that there is still a great deal of research to be
conducted in this area, despite the increasing interest.
Enthusiasts of technological advancements believe
artificial intelligence (AI) is a permanent fixture in our society, supported
by research findings and its current growth and presence in various aspects of
human life (S. Lee et al., 2022). The majority of AI
initiatives aim to achieve permanent improvement, thus increasing expectations
for its continued use. The integration of intelligent algorithms has
revolutionized digital technologies in our daily lives, particularly through
automated problem-solving processes (Raphael, 2022) and personalized
digital services (Maksimova et al.,
2021).
However, AI also raises concerns such as privacy (Hu & Min, 2023), information
security (S. Lee et al., 2020), bias and the reliability
of decision-making systems (Qiu et al., 2022;
Sun et al., 2022), issues discussed
from different critical perspectives. Among the recent AI developments are
chatbots, also known as conversational robots, agents, or personalized
assistants, which interact and "talk through text" with human users.
They have been used mainly in customer service systems (Antonio et al.,
2022), personal and home
assistance, e-commerce, marketing and business management (Reis et al., 2022), transportation and
logistics (Aksyonov et al.,
2021) and
citizen-government interaction.
Chatbots are based on natural language models, which
assimilate human language structure, identify patterns, make predictions, and
generate conversational responses through training with large data amounts and
algorithms (C.-C. Lin et al.,
2023). There are two
types: "open" or general, available to the public and answering
various topics; and "closed" or specific, designed for particular
fields like customer service or patient care (Wilson &
Marasoiu, 2022). Their creation
requires substantial information to answer diverse user questions and constant
updating and training to keep responses relevant, involving significant time
and cost (Al-Tuama &
Nasrawi, 2022).
In education, chatbot use is emerging and generating
interest though academic publications are minimal due to novelty (Bailey &
Almusharraf, 2021). Initial literature
shows positive expectations, focusing on identifying AI developments'
intentionality and application in university courses as virtual assistants or
tutors, supporting mass or self-directed learning models (Hsu & Huang,
2022), or mediating
students' emotional regulation (Benke et al., 2020). While some skeptics
exist (Winkler &
Söllner, 2018), recent reviews have
examined chatbots for Facebook Messenger as learning support (Smutny &
Schreiberova, 2020), attempts to use
chatbots in education (Kuhail et al., 2023), generative AI
research trends in educational praxis (Bozkurt, 2023), chatbot use trends
in educational contexts (Hwang & Chang,
2023), and benefits,
opportunities, challenges, and perspectives of AI chatbots in education (Labadze et al.,
2023). However, a specific
review complementing these objectives is required to further explore the
potential benefits and suitability of natural language model advancements for
higher education.
2. Methodology
According to Carrera-Rivera (2022), conducting a
literature review facilitates the identification of specific ideas or patterns
of ideas that contribute to understanding extensive information. In this study,
the literature review process followed the phases mentioned by said researcher,
and articulated with the guidelines of PRISMA method, as illustrated in Figure
2.
Figure 1
Review method design
2.1 Review protocol design
The initial stage of the literature review involved
determining its purpose, which aimed to identify the key transformations in
teaching practices resulting from the increased utilization of chatbots and
other developments in artificial intelligence. To guide this review, a research
question was formulated: "What are the effects of chatbot implementation
on teacher practice?" Following the research question, the next step
involved selecting appropriate sources of information.
Scopus, a comprehensive journal database known for its
rigorous review and editorial processes, was chosen due to its broad coverage
and diverse range of journals, taking into account Scielo and DOAJ as
complementary databases. According to Pranckutė (2021), these databases
have high academic and scientific recognition due to the rigor of their blind
peer review processes and have very strict editorial policies, ensuring good
quality of the sources to be reviewed. On the other hand, this set of databases
provides broad thematic coverage and a high number of high-impact journals to
work with. Finally, especially Scopus, offers reviewers a set of data analysis
tools that are very useful in the initial stages of the review.
To address the review question, a keyword string was
applied in Scopus, comprising the following terms:
TITLE-ABS-KEY ("teacher practice" OR "teaching
practice" OR
teaching) AND (chatbots OR "artificial
intelligence") AND (LIMIT TO SUBJAREA,“SOCI“).
2.2 Literature search and
study selection
In this phase, three characteristic processes of the
PRISMA method were applied: identification, screening, and eligibility.
The initial search yielded a total of 2683 documents
after social science filtering (Scopus=2442, Scielo=25, and DOAJ=216). To
ensure a suitable sample for further analysis, a probabilistic representative
sample of 337 documents was calculated, with a 95% confidence level and a 5%
margin of error.
For the calculation of this sample S, the following
formula was applied, where N = the size of the initial set of documents, e =
the margin of error, and z = z score, which is defined as the number of
standard deviations that a given proportion deviates from the mean.
S=
Finally, 85 duplicated articles (repeated in the
databases) were eliminated.
As part of the eligibility step, an abstracting process
was conducted, in which the following inclusion/exclusion criteria were applied
to ensure the relevance and quality of the included studies. (1) they directly
addressed the use of chatbots or artificial intelligence in educational
contexts from a pedagogical perspective; and (2) they presented empirical data
supporting the reported findings. Additionally, articles had to be published in
peer-reviewed journals indexed between 2015 and 2023 and written in English or
Spanish. As exclusion criteria, duplicate studies, theoretical reviews without
empirical data, and works that did not offer clear contributions to the
review’s objective were discarded. These criteria ensured a pertinent,
up-to-date, and methodologically sound research corpus. The documents that met
these criteria comprised the set of documents subjected to in-depth reading
(n=155).
To ensure the rigor of this review, a systematic
evaluation of the quality of the included studies was conducted. Each article
was assessed based on thematic relevance, applied methodology, and the
robustness of the reported findings. The evaluation was focused on parameters
such as clarity of objectives, validity of methods, reliability of data
collection and analysis, and well-supported conclusions. This evaluation
allowed prioritization of studies that provided significant and well-documented
contributions to analyzing the effects of chatbots in education.
2.3 Data Extraction and
Analysis
The data extraction phase involved meticulously
reading each selected article and recording relevant information in a
documentation matrix, where the data was systematically analyzed. The data
analysis followed a mixed approach combining qualitative (grouping and
categorization) and quantitative techniques (analysis of frequencies or
co-occurrences). Initially, open coding was applied to identify emerging
concepts and patterns, which were then organized into main thematic categories
through inductive analysis. Subsequently, axial coding was employed to
establish relationships between categories, enabling a deeper understanding of
the studied phenomena.
The analysis of co-occurrences involved examining how
often specific themes or keywords appeared together within the same article or
section. A co-occurrence matrix was created to quantify and visualize the
relationships between different concepts. For instance, themes such as
"pedagogical transformation, "personalized learning," and
"student engagement" were frequently linked, indicating a strong
interrelation in the context of AI applications in education. This step was
facilitated by using specialized software for text analysis, ensuring precision
and consistency. Finally, the results of the frequency and co-occurrence
analysis were synthesized into a visual representation, such as heatmaps or
network diagrams, to highlight the most significant connections and patterns.
The final phase of the review encompassed
synthesizing, interpreting, and compiling the results into a coherent text. The
findings were structured according to the IMRaD (Introduction, Methods,
Results, and Discussion) format, facilitating a comprehensive understanding of
the research outcomes. In this stage, both qualitative and quantitative
analyses were performed, ensuring a rigorous examination of the collected data.
The researchers meticulously analyzed the data for accuracy and relevance,
extracting key insights and trends. Subsequently, the synthesized findings were
interpreted to provide a deeper understanding of the research subject. Finally,
the researchers organized and compiled the results into a cohesive text,
presenting the methodology, results, and subsequent discussions systematically
and logically.
3. Analysis and
results
3.1 Main effects of chatbot
implementation on teacher work
The Figure 3 provides a visual representation of the
key themes and concepts emerging from the analysis of the integration of
artificial intelligence and chatbots in education.
Figure 3
Key themes and concepts
related to results
One of the first issues identified in the literature
regarding AI-based tools is the emotional response to their implementation.
While 35.6% of studies express a hopeful and positive outlook on chatbots in
education, 28.2% reflect feelings of risk and distrust, often echoed in the
media. Aoun (2017) highlights that AI
and robotics have outperformed humans in specific tasks, prompting reflection
on roles where humans excel, such as fostering creativity and adaptability, and
discouraging outdated training practices. This perspective is supported by López
Regalado et al. (2024) and Villegas-José and Delgado-García (2024).
As documented in 67.3% of reviewed studies, chatbots
are increasingly used in education for tasks such as administrative support and
dropout prediction. They also assist teaching by addressing student doubts and
simplifying complex topics (K.-C. Lin et al.,
2023). Moreover, 28.7% of
articles emphasize that automating repetitive tasks for teachers can improve
teaching quality by freeing time for course design and personalized feedback (Su & Yang, 2023). Chatbots also
encourage student participation by providing a pressure-free environment for
inquiries.
In massive education models like MOOCs, chatbots play
a complementary role, simulating teacher-student interactions otherwise limited
by scale. Although only 7.4% of studies explore chatbots in MOOCs, their
relevance in digitally mediated learning is notable, as noted by Li (2022) and Bachiri and
Mouncif (2023). These findings underscore the dual potential and limitations of chatbots
in education, requiring further exploration.
3.2 Disruption-related results
According to Aoun (2017), from time to time
technological developments appear on the human scene with sufficient
capabilities to radically transform life in all its dimensions. It happened
with industrialization and mechanization coming from steam technology, with
electricity, with the Internet, and now, with robotics and artificial
intelligence. In this regard, those who have followed up on these phenomena
agree that the arrival of these technologies, in terms of work and professional
spaces, always means that some are lost, and some are transformed or emerge (Mesquita et al.,
2021).
Such reflection, taken to the subject that has been
exposed in this text, puts us in a position to ask ourselves: Because of
artificial intelligence... What issues of a teacher will be lost? What should
be transformed? What new roles should the teacher assume? In other words, what
would a teacher do better than a robot or an artificial intelligence system?
Therefore, throughout some of the results of this
review, we want to address possible answers to these questions, which become
essential for teachers´ relevance within an educational system that is taking
increasingly decisive steps toward the structural incorporation of
transformative technologies such as artificial intelligence. From this point of
view, we have organized the following results.
The Figure 4 provides a visual representation of the
disruption-related results.
Figure 4
Disruption-related results
3.2.1 About what
teachers will miss out on due to chatbots
By acknowledging that the scope of pedagogy
encompasses education in its entirety, it becomes evident that many of the
challenges commonly encountered in educational practices are likely to be
impacted by the emergence of robust digital technologies such as Artificial
Intelligence. Consequently, a pedagogical perspective must be employed to
analyze and understand these changes in a natural and composed manner. This
will aid in the adaptation to new discourses and professional practices of
teachers. Here are some issues found in the literature about that:
Loss #1: The
teacher's role as a transmitter of information.
Since the mid-1990s, concerns have emerged about
digital technologies threatening teachers' jobs. Literature (72.8%) highlights
the growing role of AI in education, providing students access to vast
information in diverse formats and fueling tensions between teachers and
chatbots (Malik et al., 2021;
Safadel et al., 2023). However, the idea
that chatbots will eliminate teachers' roles as information transmitters is
debated.
Chatbots currently lack the ability to recognize
individual student characteristics, limiting their capacity to adapt to diverse
learning needs. In contrast, teachers excel at personalizing instruction,
offering feedback, and providing emotional support—roles that AI cannot fully
replicate (Meng & Dai,
2021).These human-centered
elements remain central to effective education.
Nonetheless, chatbots could replace the informational
role of some teachers by providing precise and readily accessible content. This
is more likely in contexts where teaching focuses solely on delivering
information. However, in regions with limited digital infrastructure, teachers
remain essential as content transmitters. This underscores that while AI can
supplement education, its impact is shaped by context, infrastructure, and
teaching approaches. The relationship between AI and educators should focus on
complementarity rather than replacement, ensuring that human-centric teaching
continues to enrich educational experiences.
Loss #2: Homework exclusively
related to reporting data or information.
Chatbots provide students with immediate access to
information and answers, eliminating the need to spend hours searching across
various sources. As noted in 17.4% of reviewed articles, this capability allows
students to quickly obtain necessary information through text chats,
streamlining tasks that previously relied on extensive data gathering.
Consequently, assignments focused on reporting information have become less
relevant, enabling both students and teachers to focus more on tasks that
involve analyzing and understanding the acquired information, as noted by Fidan
and Gencel (2022) and Malik et al. (2021).
This shift necessitates a transformation in the design
of homework and educational activities. Assignments should aim to strengthen
students' abilities rather than diminish their learning opportunities due to
over-reliance on chatbots. Moreover, higher education institutions should
consider adopting tools for similarity verification and detecting
machine-generated writing. This would introduce scenarios where artificial
intelligence is used to identify AI-generated content.
However, the emphasis in evaluation should shift away
from the production of text itself. Instead, the focus must be on students'
ability to comprehend, analyze, and engage with the text. This ensures that
educational assessments prioritize critical thinking and understanding over
rote production, aligning learning objectives with the evolving use of AI
technologies in education.
Loss #3: Evaluation
for all equally based solely on memory.
In consideration of the above, a third issue was
extracted from the literature reviewed (8,2%) that focused on the assessment of
learning. So, when a student relies on chatbots to report information, the
evaluation mechanisms focused on said processes would no longer make sense. For
this reason, in the evaluation framework, it will be important to resort to
other ways of assessing learning results, such as discussions, debates,
projects, portfolios, or practices that, in addition to allowing verification
of the authenticity of the student's intellectual production against the
possibility of using chatbots, allow the teacher to identify their performance
directly. Some research that addressed these topics are Ledwos et al. (2022) and Chou (2023).
This is nothing more than the claim of formative
assessment over the summative so that through it the various possibilities of
AI are used as part of learning assessment activities.
On the other hand, involving chatbots and other
developments based on artificial intelligence in the evaluation of learning
could lead to the implementation of evaluation processes where different
evaluation methods and instruments are applied to different students. Perhaps
we are at the beginning of the fall of the homogenized and standardized
evaluation.
3.2.2 About transformations that will affect
teachers due to chatbots
Some of the issues that will tend to be transformed
due to the progressive use of chatbots in education are related to what
Zambrano (2005) points out about
Pedagogy, in terms of conceiving it as a discourse on relationships between
teachers, students, the school and social environment and the forms of the
orientation of knowledge that take place.
Transformation # 1:
About control over the intentionality and orientation of educational content.
Considering the above, a few percentage of the
articles reviewed (5,8%), report that the use of chatbots in education has to
do with the transfer of the monopoly of control that teachers and the school
institution have had so far over the intentionality and orientation of the
students' learning content. Historically, students receive during their school
life a set of structured knowledge in the form of curricular proposals, which
someone has estimated correspond to what should be learned. So, with what intention
has the curriculum been organized like this? Is it okay for one vague person to
determine what another person should learn? Who decides this? Certainly not the
student. This is something that has not been questioned enough and that is
accepted as part of the current paradigm of education and that, due to the use
of artificial intelligence developments in education, is beginning to be
questioned. Some of the above can be found in Farhi et al. (2022) or Chassignol et al.
(2018).
In this sense, the chatbot can offer a personalized
learning experience, adapted to the needs and preferences of the student,
allowing them to explore and build their knowledge in a more autonomous way (Srimathi &
Krishnamoorthy, 2019). However, this
paradigm shift also entails certain challenges and risks, one of the main ones
being maintaining a high level of quality and consistency in content and
learning orientation, since the chatbot cannot always guarantee that students
receive the correct and relevant information.
Transformation # 2:
Who will be in charge of the didactic transposition?
A small percentage (4.2%) of reviewed documents
address educational content creation processes, specifically focusing on
didactic transposition. This concept, developed in the 20th century, describes
the transformation of scientific knowledge into teachable material and
ultimately into knowledge that students can understand and learn (Chevallard, 1998). This
"translation" process ensures content aligns with students' cognitive
development, language, and prior knowledge, traditionally managed by teachers
or subject-matter experts.
Generative AI is now playing a role in didactic
transposition, as natural language models are designed not only to provide
answers but also to simplify and explain scientific knowledge in accessible
terms. This linguistic capability positions AI as a valuable tool in
harmonizing complex concepts with everyday language.
Moreover, AI systems can be trained to identify
individual learning styles, limitations, and abilities, allowing the
transposition process to cater more closely to each student's needs. This
enables a more personalized approach to learning, complementing teachers’ roles
in content adaptation. By supporting these processes, AI has the potential to
enhance educational content delivery, ensuring accessibility and relevance.
Examples of such AI applications in content creation are discussed by Ohanian (2019) and Ako-Nai et al. (2022), demonstrating its
growing influence in educational innovation.
Transformation# 3:
The didactic contract.
Finally, the last few of the articles reviewed (3,7%)
refer to potential changes in teacher-student relationships. Regarding this, in
the context of the use of chatbots in education, the "didactic
contract" becomes an important concept related to such relationships, with
big and complex challenges ahead.
Didactic Contract refers to the tacit agreement
between the teacher and the student about what is expected to happen in the
classroom and how learning will take place. This contract establishes the rules
and expectations for learning and can influence how chatbots are used in the
classroom (Caldeborg et al.,
2019).
In the context of the use of chatbots in education, the
didactic contract can be challenged by the introduction of new technological
tools. For example, students may expect a more personalized interaction with
the chatbot, which may require the teacher to adapt their teaching approach and
strategies to meet those needs. Research related to changes in classroom
relationships can be found in Garito (1991) or Lo et al. (2021).
4. Discussion and
Conclusions
The deployment of AI chatbots in educational settings
presents a multifaceted issue that demands profound pedagogical examination.
The use of chatbots and AI tools in education introduces significant changes in
pedagogical practices. Chatbots can automate repetitive tasks, such as
answering common questions, allowing teachers to focus on higher-value
activities like lesson design and personalized student support. This shift can
foster active learning and collaboration in the classroom. However, these tools
require teachers to adapt their roles, acting as facilitators and mediators of
responsible technology use. Chatbots promote self-directed learning but demand
critical skills to evaluate information. Additionally, assessments must
emphasize critical thinking and creativity rather than memory-based tasks. In
this regard, the teacher-student relationship remains crucial. While chatbots
personalize learning, human interaction fosters empathy, motivation, and
emotional support. Effective AI integration must align with pedagogical
principles that prioritize holistic student development.
AI-driven chatbots hold significant promise in
automating teaching tasks, offering efficiencies and accessibility previously
unattainable. Nevertheless, they cannot fully replicate the unique qualities of
human interaction essential to education, such as empathy, emotional
intelligence, adaptability, and the ability to inspire and motivate learners.
Indeed, these deeply human attributes transcend mere information transmission
and often resist replication by even the most advanced algorithms.
Therefore, integrating AI chatbots into education
requires a critical assessment of their strengths and limitations from a
pedagogical perspective. For instance, research should identify areas where
chatbots excel, such as automating repetitive tasks, while highlighting their
shortcomings, particularly in fostering meaningful human connections. By doing
so, educators can leverage chatbots in tasks where automation is beneficial,
freeing instructional time for activities that demand the irreplaceable human touch.
In this context, the interaction between generative
chatbots and teachers represents a dynamic relationship where both must
complement each other's strengths to create an effective educational system.
Consequently, future studies should examine chatbot-student interaction designs
and explore the impact of chatbot personality and location on learning outcomes
and satisfaction. Furthermore, the rapid evolution of AI in education
necessitates mechanisms to maximize its potential while addressing challenges
such as emotional intelligence and ethical use.
As tools like ChatGPT gain prominence, it becomes
evident that guidelines for their responsible adoption are critical (Tlili et al., 2023). Thus, collaboration between educators,
instructional designers, researchers, and AI developers is essential to
establish pedagogical principles that balance technological innovation with the
preservation of human elements. Ultimately, by achieving this balance, emerging
technologies can promote improved learning experiences and vital life skills,
such as self-regulation, ensuring that AI complements rather than replaces the
invaluable role of human educators (Bozkurt, 2023).
4.1 Limitations and
Recommendations
This review, while comprehensive, has limitations.
Most studies analyzed come from specific, well-resourced educational contexts,
limiting generalization to environments with fewer technological resources or
differing cultural attitudes toward AI. Besides, the focus on recent studies
reflects an evolving landscape, but the long-term impacts of chatbots remain
underexplored. Additionally, methodological inconsistencies across studies make
direct comparisons challenging. Finally, while frequency and co-occurrence
analysis identified key trends, it may overlook deeper nuances. Future research
should include qualitative methods, such as case studies, to better understand
the contextual and subjective effects of chatbots on education.
On the other hand, to optimize the integration of
chatbots in education, institutions should adopt a balanced approach that
combines technological innovation with robust pedagogical principles. Teachers
should receive training on effectively leveraging chatbots to complement, not
replace, their instructional practices. Curricula must be updated to emphasize
critical thinking, creativity, and digital literacy, enabling students to
navigate AI-enhanced learning environments responsibly. Developers should collaborate
with educators to design chatbots tailored to diverse educational contexts,
ensuring inclusivity and adaptability. Additionally, further research is needed
to explore long-term impacts, particularly on student engagement and
teacher-student dynamics, while addressing ethical concerns such as data
privacy and bias.
Author´s Contribution
All the authors participated equally in the following
processes according to the CRediT Taxonomy: Conceptualization, data curation
and formal analysis, research and methodological design, writing of the
original draft and its final review and editing. In addition, Andrés Chiappe is
the corresponding author.
Acknowledgments
We thank the Universidad de La Sabana (Group
Technologies for Academia – Proventus (Project EDUPHD-20-2022), Fundación
Universitaria Navarra - Uninavarra and Universidad Católica de la Santísima
Concepción, for the support received in the preparation of this article.
References
Ako-Nai, F., De La Cal Marin,
E., & Tan, Q. (2022). Artificial Intelligence Decision and Validation
Powered Smart Contract for Open Learning Content Creation. In J. Prieto, A.
Partida, P. Leitão, & A. Pinto (Eds.), Blockchain and Applications
(Vol. 320, pp. 359–362). Springer International Publishing.
https://doi.org/10.1007/978-3-030-86162-9_37
Aksyonov, K. A., Ziomkovskaya,
P. E., Danwitch, D., Aksyonova,
O. P., & Aksyonova, E. K. (2021). Development of
a text analysis agent for a logistics company’s Q&A system. Journal of
Physics: Conference Series, 2134(1), 012021.
https://doi.org/10.1088/1742-6596/2134/1/012021
Al-Tuama, A. T., & Nasrawi, D. A. (2022). A Survey on the Impact of Chatbots
on Marketing Activities. 2022 13th International Conference on Computing
Communication and Networking Technologies (ICCCNT), 1–7.
https://doi.org/10.1109/ICCCNT54827.2022.9984635
Antonio, R., Tyandra, N., Nusantara, L. T., Anderies, &
Agung Santoso Gunawan, A. (2022). Study Literature
Review: Discovering the Effect of Chatbot Implementation in E-commerce Customer
Service System Towards Customer Satisfaction. 2022 International Seminar on
Application for Technology of Information and Communication (iSemantic), 296–301.
https://doi.org/10.1109/iSemantic55962.2022.9920434
Aoun, J. E. (2017). Robot-Proof. The MIT Press.
https://mitpress.mit.edu/books/robot-proof
Bachiri, Y.-A., & Mouncif,
H. (2023). Artificial Intelligence System in Aid of Pedagogical Engineering for
Knowledge Assessment on MOOC Platforms: Open EdX and Moodle. International
Journal of Emerging Technologies in Learning (iJET),
18(05), 144–160. https://doi.org/10.3991/ijet.v18i05.36589
Bailey, D., & Almusharraf,
N. (2021). Investigating the Effect of Chatbot-to-User Questions and Directives
on Student Participation. 2021 1st International Conference on Artificial Intelligence
and Data Analytics (CAIDA), 85–90.
https://doi.org/10.1109/CAIDA51941.2021.9425208
Benke, I., Knierim, M. T., & Maedche, A. (2020).
Chatbot-based Emotion Management for Distributed Teams: A Participatory Design
Study. Proceedings of the ACM on Human-Computer Interaction, 4(CSCW2),
1–30. https://doi.org/10.1145/3415189
Bozkurt, A. (2023). Unleashing the Potential of
Generative AI, Conversational Agents and Chatbots in Educational Praxis: A
Systematic Review and Bibliometric Analysis of GenAI in Education. Open
Praxis, 15(4), 261–270. https://doi.org/10.55982/openpraxis.15.4.609
Caldeborg, A., Maivorsdotter, N., & Öhman, M. (2019). Touching the
didactic contract—A student perspective on intergenerational touch in PE. Sport,
Education and Society, 24(3), 256–268.
https://doi.org/10.1080/13573322.2017.1346600
Carrera-Rivera, A., Ochoa, W., Larrinaga, F., &
Lasa, G. (2022). How-to conduct a systematic literature review: A quick guide
for computer science research. MethodsX, 9,
101895. https://doi.org/10.1016/j.mex.2022.101895
Chassignol, M., Khoroshavin, A., Klimova, A., & Bilyatdinova,
A. (2018). Artificial Intelligence trends in education: A narrative overview. Procedia
Computer Science, 136, 16–24.
https://doi.org/10.1016/j.procs.2018.08.233
Chen, Y., Jensen, S., Albert, L. J., Gupta, S., &
Lee, T. (2023). Artificial Intelligence (AI) Student Assistants in the
Classroom: Designing Chatbots to Support Student Success. Information Systems Frontiers, 25(1),
161–182. https://doi.org/10.1007/s10796-022-10291-4
Chevallard, Y. (1998). La transposición didáctica
(Tercera Edición, Vol. 1). Aique.
Chou, T.-N. (2023). Apply an Integrated Responsible AI
Framework to Sustain the Assessment of Learning Effectiveness: Proceedings
of the 15th International Conference on Computer Supported Education,
142–149. https://doi.org/10.5220/0012058400003470
Farhi, F., Jeljeli, R.,
& Hamdi, M. E. (2022). How do Students Perceive Artificial Intelligence in
YouTube Educational Videos Selection? A Case Study of Al Ain City. International
Journal of Emerging Technologies in Learning (iJET),
17(22), 61–82. https://doi.org/10.3991/ijet.v17i22.33447
Fidan, M., & Gencel, N.
(2022). Supporting the Instructional Videos With Chatbot and Peer Feedback Mechanisms in Online
Learning: The Effects on Learning Performance and Intrinsic Motivation. Journal
of Educational Computing Research, 60(7), 1716–1741.
https://doi.org/10.1177/07356331221077901
Garito, M. A. (1991). Artificial intelligence in
education: Evolution of the teaching?learning
relationship. British Journal of Educational Technology, 22(1),
41–47. https://doi.org/10.1111/j.1467-8535.1991.tb00050.x
Hsu, H.-H., & Huang, N.-F. (2022). Xiao-Shih: A
Self-Enriched Question Answering Bot With Machine Learning
on Chinese-Based MOOCs. IEEE Transactions on Learning Technologies, 15(2),
223–237. https://doi.org/10.1109/TLT.2022.3162572
Hu, Y., & Min, H. (Kelly). (2023). The dark side
of artificial intelligence in service: The “watching-eye”
effect and privacy concerns. International Journal of Hospitality Management,
110, 103437. https://doi.org/10.1016/j.ijhm.2023.103437
Hwang, G.-J., & Chang, C.-Y. (2023). A review of
opportunities and challenges of chatbots in education. Interactive Learning
Environments, 31(7), 4099–4112.
https://doi.org/10.1080/10494820.2021.1952615
King, M. R. (2023). A Conversation on Artificial
Intelligence, Chatbots, and Plagiarism in Higher Education. Cellular and
Molecular Bioengineering, 16(1), 1–2.
https://doi.org/10.1007/s12195-022-00754-8
Kuhail, M. A., Alturki, N., Alramlawi, S., & Alhejori, K.
(2023). Interacting with educational chatbots: A systematic review. Education
and Information Technologies, 28(1), 973–1018.
https://doi.org/10.1007/s10639-022-11177-3
Labadze, L., Grigolia, M., & Machaidze, L.
(2023). Role of AI chatbots in education: Systematic
literature review. International Journal of Educational Technology in Higher
Education, 20(1), 56. https://doi.org/10.1186/s41239-023-00426-1
Ledwos, N., Mirchi, N., Yilmaz, R., Winkler-Schwartz,
A., Sawni, A., Fazlollahi,
A. M., Bissonnette, V., Bajunaid, K., Sabbagh, A. J.,
& Del Maestro, R. F. (2022). Assessment of learning curves on a simulated
neurosurgical task using metrics selected by artificial intelligence. Journal
of Neurosurgery, 137(4), 1160–1171.
https://doi.org/10.3171/2021.12.JNS211563
Lee, S., Lee, M., & Lee, S. (2022). What If
Artificial Intelligence Become Completely Ambient in Our Daily Lives? Exploring
Future Human-AI Interaction through High Fidelity Illustrations. International
Journal of Human–Computer Interaction, 1–19.
https://doi.org/10.1080/10447318.2022.2080155
Lee, S., Suk, J., Ha, H. R., Song, X. X., & Deng,
Y. (2020). Consumer’s Information Privacy and Security Concerns and Use of
Intelligent Technology. In T. Ahram, W. Karwowski, A. Vergnano,
F. Leali, & R. Taiar (Eds.), Intelligent Human
Systems Integration 2020 (Vol. 1131, pp. 1184–1189). Springer International
Publishing. https://doi.org/10.1007/978-3-030-39512-4_180
Li, H. (2022). MOOC Teaching Platform System Based on
Application of Artificial Intelligence. 2022 Second International Conference
on Advanced Technologies in Intelligent Control, Environment, Computing &
Communication Engineering (ICATIECE), 1–5.
https://doi.org/10.1109/ICATIECE56365.2022.10047595
Lin, C.-C., Huang, A. Y. Q., & Yang, S. J. H.
(2023). A Review of AI-Driven Conversational Chatbots Implementation
Methodologies and Challenges (1999–2022). Sustainability, 15(5),
4012. https://doi.org/10.3390/su15054012
Lin, K.-C., Cheng, I.-L., Huang, Y.-C., Wei, C.-W.,
Chang, W.-L., Huang, C., & Chen, N.-S. (2023). The Effects of the Badminton
Teaching–Assisted System using Electromyography and Gyroscope on Learners’
Badminton Skills. IEEE Transactions on Learning Technologies, 1–10.
https://doi.org/10.1109/TLT.2023.3292215
Lo, F., Su, F., Chen, S., Qiu, J., & Du, J.
(2021). Artificial Intelligence Aided Innovation Education Based on Multiple
Intelligence. 2021 IEEE International Conference on Artificial Intelligence,
Robotics, and Communication (ICAIRC), 12–15.
https://doi.org/10.1109/ICAIRC52191.2021.9544874
López Regalado, O., Núñez-Rojas, N., Rafael López Gil,
O., & Sánchez-Rodríguez, J. (2024). El Análisis del uso de la inteligencia
artificial en la educación universitaria: Una revisión sistemática (Analysis of the
use of artificial intelligence
in university education: a systematic review). Pixel-Bit, Revista de Medios y Educación, 70,
97-122. https://doi.org/10.12795/pixelbit.106336
Maksimova, M., Solvak, M., & Krimmer, R.
(2021). Data-Driven Personalized E-Government Services: Literature
Review and Case Study. In N. Edelmann, C. Csáki, S. Hofmann, T. J. Lampoltshammer, L. Alcaide Muñoz, P. Parycek,
G. Schwabe, & E. Tambouris (Eds.), Electronic
Participation (Vol. 12849, pp. 151–165). Springer International Publishing.
https://doi.org/10.1007/978-3-030-82824-0_12
Malik, R., Shrama, A.,
Trivedi, S., & Mishra, R. (2021). Adoption of Chatbots for Learning among
University Students: Role of Perceived Convenience and Enhanced Performance. International
Journal of Emerging Technologies in Learning (iJET),
16(18), 200. https://doi.org/10.3991/ijet.v16i18.24315
Meng, J., & Dai, Y. (Nancy). (2021). Emotional
Support from AI Chatbots: Should a Supportive Partner Self-Disclose or Not? Journal
of Computer-Mediated Communication, 26(4), 207–222.
https://doi.org/10.1093/jcmc/zmab005
Mesquita, A., Oliveira, L., & Sequeira, A. S.
(2021). Did AI Kill My Job?: Impacts
of the Fourth Industrial Revolution in Administrative Job Positions in
Portugal. In J.-É. Pelet (Ed.), Advances in
Business Information Systems and Analytics (pp. 124–146). IGI Global.
https://doi.org/10.4018/978-1-7998-3756-5.ch008
Ohanian, T. (2019). How
Artificial Intelligence and Machine Learning May Eventually Change Content
Creation Methodologies. SMPTE Motion Imaging Journal, 128(1),
33–40. https://doi.org/10.5594/JMI.2018.2876781
Pranckutė, R. (2021). Web
of Science (WoS) and Scopus: The Titans of
Bibliographic Information in Today’s Academic World. Publications, 9(1),
12. https://doi.org/10.3390/publications9010012
Qiu, C., Liang, W., Yan, Z., Li, Y., & You, Y.
(2022). Research and Application of Power Grid Fault Diagnosis and Auxiliary
Decision-making System Based on Artificial Intelligence Technology. 2022 7th
International Conference on Power and Renewable Energy (ICPRE), 492–497.
https://doi.org/10.1109/ICPRE55555.2022.9960335
Raphael, M. W. (2022). Artificial intelligence and the
situational rationality of diagnosis: Human problem‐solving and the artifacts of health and medicine. Sociology
Compass, 16(11). https://doi.org/10.1111/soc4.13047
Reis, L., Maier, C., & Weitzel, T. (2022).
Chatbots in Marketing: An In-Deep Case Study Capturing Future Perspectives of
AI in Advertising. Proceedings of the Conference on Computers and People
Research, 1–8. https://doi.org/10.1145/3510606.3550204
Safadel, P., Hwang, S.
N., & Perrin, J. M. (2023). User Acceptance of a Virtual Librarian Chatbot:
An Implementation Method Using IBM Watson Natural Language Processing in
Virtual Immersive Environment. TechTrends.
https://doi.org/10.1007/s11528-023-00881-7
Salvagno, M., Taccone,
F. S., & Gerli, A. G. (2023). Can artificial
intelligence help for scientific writing? Critical
Care, 27(1), 75. https://doi.org/10.1186/s13054-023-04380-2
Smutny, P., & Schreiberova,
P. (2020). Chatbots for learning: A review of educational chatbots for the Facebook Messenger. Computers & Education, 151,
103862. https://doi.org/10.1016/j.compedu.2020.103862
Srimathi, H., & Krishnamoorthy, A. (2019).
Personalization of student support services using chatbot. International
Journal of Scientific and Technology Research, 8(9), 1744–1747.
Scopus.
Su, J., & Yang, W. (2023). Unlocking the Power of
ChatGPT: A Framework for Applying Generative AI in Education. ECNU Review of
Education, 6(3), 355–366. https://doi.org/10.1177/20965311231168423
Sun, J., Zhang, X., & Ding, Y. (2022). Research on
trustworthiness assessment technology of intelligent decision-making system. In
S. Yang & G. Wu (Eds.), Third International Conference on Artificial
Intelligence and Electromechanical Automation (AIEA 2022) (p. 78). SPIE.
https://doi.org/10.1117/12.2646844
Surahman, E., & Wang,
T. (2022). Academic dishonesty and trustworthy assessment in online learning: A
systematic literature review. Journal of Computer Assisted Learning, 38(6),
1535–1553. https://doi.org/10.1111/jcal.12708
Tlili, A., Shehata, B., Adarkwah,
M. A., Bozkurt, A., Hickey, D. T., Huang, R., & Agyemang, B. (2023). What
if the devil is my guardian angel: ChatGPT as a case study of using chatbots in
education. Smart Learning Environments, 10(1), 15.
https://doi.org/10.1186/s40561-023-00237-x
Villegas-José, V., &
Delgado-García, M. (2024). Inteligencia artificial: Revolución educativa
innovadora en la Educación Superior. Pixel-Bit,
Revista de Medios y Educación, 71, 159-177. https://doi.org/10.12795/pixelbit.107760
Wilson, L., & Marasoiu,
M. (2022). The Development and Use of Chatbots in Public Health:
Scoping Review. JMIR Human Factors, 9(4), e35882.
https://doi.org/10.2196/35882
Winkler, R., & Söllner,
M. (2018). Unleashing the potential of chatbots in education: A
state-of-the-art analysis. Academy of Management Annual Meeting (AOM).
Zambrano, A. (2005). Didáctica, pedagogía y saber:
Aportes desde las ciencias de la educación. Editorial Magisterio.