Referencia de Articulo 1:

Zhu, M., Izhar, N. A., & Xervaser, C. R. (2025). Educators’ perspectives on DeepSeek in
ELT: A qualitative case study of pedagogical potentials and pitfalls in Chinese higher
education [Perspectivas de los educadores sobre DeepSeek en la enseñanza del inglés:
Estudio de caso cualitativo sobre potenciales y riesgos pedagógicos en la educación
superior china]. Journal of Information Technology Education: Research, 24, 1–25.
[Link]
Conclusiones de Articulo 1:
“Integrating DeepSeek into ELT in higher education plays an important role across
pedagogy, curriculum, and technology. DeepSeek enhances teaching practices by
streamlining lesson planning, increasing efficiency, and enabling real-time instructional
adjustments and feedback. Rather than replacing human expertise, it supports personalized,
interactive instruction while maintaining educator autonomy. However, these potentials could
be realized through ongoing professional development and institutional support that enables
educators to use AI tools critically and effectively” (p.19)
“KEY FINDINGS This study empirically explores the use of DeepSeek in teaching English in
higher education from the perspective of educators’ experiences, filling a gap in the field of
ELT and learning research on educators’ subjective experiences with this AI tool. The first
theme, Transformative Pedagogical Support, is about the advantages of DeepSeek. In the
Before phase, while the tool demonstrated benefits in enhancing preparatory efficiency
through structured content generation and resource enrichment, its classroom application
revealed critical limitations, including system instability, contextual misalignment of outputs,
and risks of student over-reliance that could undermine long-term language development.
Notably, educators emerged not as passive adopters but as active mediators of technology,
strategically adapting its functionalities while maintaining pedagogical authority. The study
highlights the necessity of comprehensive training programs that address both technical
proficiency and critical AI literacy while suggesting that institutional policies must evolve to
support educators in navigating the complexity of technological affordances and pedagogy.
Future research should investigate scalable implementation models across diverse
educational contexts to further explore the conditions under which open-source AI tools can
most effectively empower language instruction without compromising educational quality or
student autonomy. In summary, DeepSeek demonstrates significant instructional support
value in ELT, but its functionality is limited, while also promoting a positive transformation of
the educator’s role. IMPLICATIONS The theoretical significance of this study lies in its
provision of a localized empirical basis for the application of DeepSeek AI tools in ELT in
higher education from the educator's core perspective, thereby enriching the theoretical
exploration of the interaction between technology and educators in the field of
“AI+Education.” In addition, it emphasizes the key role of educators’ initiative in the process
of technology-enabled teaching and learning. For example, the functional value of AI tools in
improving efficiency needs to be transformed into teaching effectiveness through educators’
improvement of their own autonomy in curriculum design and their ability to respond in real
time. Educators’ selective use of AI tools (modifying tool suggestions to fit ideological
teaching requirements) revealed that educators are not passive recipients of technology but
define the boundaries of technology and lead the direction of its application through
professional judgement, which provides a new case for the extension of the theory of
educator agency in the digital age. The practical significance of this study is evident in the
results, which provide clear guidelines for university English educators on the effective use
of DeepSeek in optimizing their teaching practice. Before teaching, they can use the tool to
expand their ideas of lesson design, and the need to anchor teaching objectives to avoid
deviating from the core. During class, they can use the tool to enhance classroom interaction
and guidance, while also controlling the pace of the lesson according to the students’
responses to ensure the appropriateness of the lessons. After class, the tools can be used to
improve the efficiency of feedback, and attention to individual differences in students should
be added to strengthen learning outcomes. For educational institutions, the study provides a
practical reference for promoting the integration of AI and teaching to address the limitations
of the tools (such as system instability and insufficient content adaptability). We should
perhaps offer feedback to the developers of the tools to enhance the practicality of usage in
our contexts. Also, to address the demand for transformation of educators’ roles, we need to
design targeted training (such as curricular innovation and data interpretation ability) to help
educators master these tools. This synergistic path of educator practice and institutional
support could promote AI tools from simple applications to in-depth integration and
effectively meet the actual needs of teaching” (p.20)
“FUTURE DIRECTION Further research should be undertaken to investigate the different
types of colleges and universities in more subdivided scenarios of English teaching,
exploring in-depth the adaptation strategies of AI tools and diversified teaching needs.
Additionally, to further expand the breadth and depth of the research, it is necessary to
explore in-depth how to develop a more effective educator training system to enhance the
ability and quality of educators to utilize AI technology in response to the rapid development
of technological advancements. In terms of practical promotion, it is necessary to explore
how to establish an effective incentive mechanism to encourage educators to proactively
integrate AI technology into daily teaching and realize the deep integration of technology and
teaching, giving a full empowering effect of AI technology on education, and laying a solid
foundation for the cultivation of innovative talents to meet the needs of future society” (p.21).

Referencia de Articulo 2:
Muñoz Martínez, C., Roger-Monzo, V., & Castelló Sirvent, F. (2025). IA generativa y
pensamiento crítico en la educación universitaria a distancia: Desafíos y oportunidades.
RIED – Revista Iberoamericana de Educación a Distancia, 28(2), 1–20.
[Link]

Conclusiones de Articulo 2:
“La investigación sobre el desarrollo del pensamiento crítico en contextos de uso
generalizado de IAG sigue siendo limitada y fragmentaria. Este estudio exploratorio
constituye un primer esfuerzo por identificar los factores que inciden en su desarrollo,
analizando barreras, aceleradores, alternativas tecnológicas, retos sociales y
consecuencias de su integración. Los consensos alcanzados y los vectores de actuación
propuestos pueden servir como base para el diseño de instrumentos de medición, como
encuestas dirigidas a docentes universitarios. Contar con una muestra suficientemente
amplia y diversificada permitirá contrastar las percepciones expertas con datos
empíricos, fortaleciendo la validez externa y fiabilidad de los hallazgos. Esta
triangulación permitirá evaluar la coherencia entre tendencias cualitativas y
mediciones cuantitativas, identificando posibles discrepancias o convergencias en la
relación entre IAG y pensamiento crítico, y proporcionando un marco más sólido para la
toma de decisiones en el diseño de políticas pedagógicas y estrategias de formación
[Link] este estudio se ha desarrollado en universidades a distancia dentro
del Espacio Europeo de Educación Superior (EEES), lo que podría restringir la
generalización de los resultados a otros sistemas con normativas y estructuras
organizativas distintas, los consensos alcanzados y las líneas de actuación propuestas
muestran una alta capacidad de adaptación a distintos entornos educativos digitales,
favoreciendo su aplicabilidad en contextos con dinámicas de aprendizaje [Link] bien
esta investigación ha centrado el análisis en el rol del profesorado en la
implementación de la IAG, futuros trabajos deberían ampliar el espectro de actores
involucrados, incluyendo la perspectiva de estudiantes, responsables de políticas
educativas y representantes del sector productivo. Incorporar estos perfiles permitiría un
enfoque más holístico y multidimensional, facilitando una comprensión más completa
de los desafíos y oportunidades que la IAG plantea en el desarrollo del pensamiento
crítico. Esto enriquecería el debate académico y fortalecería el diseño de estrategias
pedagógicas más contextualizadas y ajustadas a las necesidades del ecosistema
educativo” (p.16)
“La IAG puede hacer el aprendizaje más inmersivo, dinámico y personalizado,
mejorando el rendimiento académico, la motivación y la autorregulación (Huang et al., 2023;
Yuan y Liu, 2025). No obstante, sin una mediación adecuada, su uso podría generar
una dependenciaacrítica, llevando a los estudiantes a aceptar respuestas
automatizadas sin analizarlas reflexivamente, lo que limitaría su capacidad de
cuestionamiento (Chng et al., 2023). Para mitigar este riesgo, es fundamental
equilibrar la integración tecnológica con la intervención docente, asegurando que la
IAG no reemplace el proceso cognitivo del estudiante, sino que lo potencie a través de
estrategias estructuradas y [Link] engagement en educación a distancia no solo
depende del acceso a herramientas innovadoras, sino también de la calidad de la
interacción docentes-estudiantes, un factor determinante para la retención del
conocimiento y la satisfacción académica (Bae et al., 2020; Hoi y Le Hang, 2021).
Por ello, la planificación pedagógica debe integrar la IAG sin desplazar la labor
docente, alineándose conobjetivos que promuevan la validación de información, la
identificación de sesgos algorítmicos y la reflexión sobre el impacto de la IA en el
aprendizaje (Martín y López, 2023).Para una integración efectiva, los docentes requieren de
respaldo institucional. Las instituciones deben establecer directrices que regulen su
uso asegurando una implementación ética, accesible y alineada con estándares
pedagógicos que estimulen el pensamiento crítico. Asimismo, la formación docente en
alfabetización digital e IA resulta imprescindible para diseñar estrategias de evaluación
que fomenten un aprendizaje auténtico. En entornos de aprendizaje a distancia, donde el
ensayo ha sido la principal herramienta de evaluación, contrastar respuestas generadas
por IA con fuentes verificadas, diseñar promptsestratégicos y analizar críticamente los
outputsalgorítmicos pueden fomentar una interacción más reflexiva con la
tecnologí[Link] estudios empíricos deberían analizar cómo las intervenciones educativas
con herramientas de IA impactan en el pensamiento crítico y en las dimensiones
afectiva y conductual del engagement estudiantil. Para ello, resultaría útil combinar
instrumentos de evaluación de habilidades cognitivas con autoinformes, análisis de
interacciones y métricas cuantitativas extraídas de plataformas virtuales. Además,
estudios longitudinales podrían examinar su efecto a medio plazo, identificando su
influencia en la reducción del abandono, la mitigación de la desconexión y el
aislamiento académico, así como su contribución al desarrollo de la autonomía, la
autorregulación del aprendizaje y la motivación estudiantil.” (p.17)
Referencia de Articulo 3:
Weigelt, H., Segev, E., Kurtz, G., Kahana, O., & Fogel, N. R. (2026). Enhancing students’
critical thinking literacy in a generative AI context: Eye movement patterns of deepfake
detection [Mejora de la alfabetización en pensamiento crítico de los estudiantes en un
contexto de IA generativa: Patrones de movimiento ocular en la detección de deepfakes].
Computers & Education, 208, 105529. [Link]

Conclusiones de Articulo 3:
“This exploratory study examined, using eye-tracking technology, the gaze patterns of
guided exposure to deepfake and authentic portrait images in comparison to unguided
exposure to assess their influence on the acquisition and application of critical thinking
literacy among first-year Bachler students. Eye-tracking represents a central methodological
contribution of this study, as it provides objective indicators of participants’ cognitive
processing during the intervention. Prior research has shown that eye-movements can
distinguish learning strategies, support analysis of problem-solving, and capture local versus
global visual attention (Catrysse et al., 2018; Schreiter & Vogel, 2023; Soares et al., 2021).
By applying this method to the context of deepfake detection, the study not only advances
understanding of how guided versus unguided exposure shapes gaze behaviour but also
demonstrates the broader potential of eye-tracking for investigating how educational
interventions influence students’ critical thinking processes. The results, while preliminary,
show a significant difference in visual attention patterns between the experimental and
control groups. These initial findings suggest that targeted training can alter the cognitive
processing strategies students employ when evaluating GenAI-generated content.
Specifically, the experimental group, which received critical thinking intervention, exhibited a
wider distribution of fixations compared to the control group (H1 accepted, H2 partially
accepted). As in previous studies (Cheetham et al., 2013; Gupta et al., 2020; Huang et al.,
2024; Sano et al., 2024; Wohler ¨ et al., 2021), both groups spend similar total fixation
durations across the main facial features of eye, nose and mouth. Focusing on these
features while examining a face is mainly explained by psychological and social aspects,
since these features are critical for conveying emotional expressions, social signals, and
identity recognition (Porras & Louwerse, 2025). However, the finding that the experiment
group significantly expanded their visual attention to peripheral areas (higher total fixation
duration of background AOI’s), may indicate a shift from intuitive visual examination strategy
to a more comprehensive evaluation approach, emphasizing the interventions impact. This
finding is particularly significant given the challenges identified in previous deepfake
detection research (Gupta et al., 2020; Huang et al., 2024). Huang et al. (2024) found that
an image’s background plays a key role in how viewers judge its source. Yet, most visual
attention typically focuses on the central facial features - the eyes, mouth, and nose
(Fearington et al., 2025). Building on this, the current study suggests that the intervention
enhanced students’ critical thinking literacy and encouraged them to broaden their visual
attention to image areas that support more accurate deepfake detection. The significantly
better overall detection accuracy demonstrated by the experiment group in comparison to
the control group (H3 accepted) further supports the intervention’s positive impact on
students’ practical skills in detecting deepfakes. Nevertheless, it is crucial to note that both
groups still exhibited relatively low accuracy for individual images. This outcome aligns with
the literature, which consistently reports the inherent difficulty humans face in distinguishing
between GenAI-generated content and authentic media, especially as GenAI technology
rapidly advances (Huang et al., 2024; Nightingale & Farid, 2022). While the intervention
improved overall accuracy, the persistent challenge at the individual image level
underscores the sophistication of current deepfake algorithms. This suggests that even with
explicit guidance, the visual cues for detection can be subtle, inconsistent, or even absent in
highly convincing deepfakes. The results align with those of Bray et al. (2023), who found
that while an intervention increases accuracy, the overall improvement may not always be
substantial enough to overcome the fundamental difficulty posed by highly realistic synthetic
media. Diel et al. (2024) suggested combining several intervention strategies, such as
feedback training and raising awareness, to improve detection. The implication for education
is that, while training is beneficial, it must be continuous, diverse and adaptive to the
evolving visual capabilities of GenAI. The findings revealed that the control group tended to
classify the images as authentic compared to the experimental group. The control group’s
tendency to classify more images as authentic, even when they were deepfakes, aligns with
the documented AI hyperrealism bias., e.g., a higher tendency to choose GenAI images of
faces as authentic humans (Nightingale & Farid, 2022). Participants” (.p12)
“in the experimental group demonstrated a greater propensity to classify images as
deepfakes compared to the control group (H4 rejected). This result is consistent with findings
by Chen et al. (2025) and Rehman et al. (2025), who also observed that AI detection training
increases participants’ tendency to identify images as fake. These findings carry critical
implications for educational practice. While training can improve visual sampling strategies
and help mitigate biases such as hyperrealism, it may also introduce a new bias toward
over-identifying deepfake. To translate improved information-gathering into more accurate
judgments, additional pedagogical elements are needed. By providing specific principles and
strategies for deepfake detection, the intervention equipped students with cognitive tools to
critically scrutinize images rather than defaulting to an assumption of authenticity. This is a
critical outcome for critical thinking literacy in a GenAI-saturated information environment, as
it demonstrates the potential to mitigate inherent biases that can lead to misinterpretations of
digital content. Contrary to the hypothesis, there was no significant difference between the
experimental and control groups in their perceived success of identification (H5 rejected).
Both groups similarly overestimated their performance, with self-assessed estimates
significantly higher than their actual accuracy. This finding aligns with the existing literature
(Kobis ¨ et al., 2021), which suggests that individuals often struggle to accurately gauge their
performance on complex tasks, especially when guessing is involved (Vuorre & Metcalfe,
2022). This persistent overestimation, even in the experiment group, suggests that while the
intervention improved actual detection skills and altered visual gaze patterns, it did not
necessarily enhance metacognitive awareness of those improvements. It is possible that the
cognitive load associated with the deepfake detection task made it challenging for
participants to accurately monitor their performance. Another explanation, also suggested by
Kobis ¨ et al. (2021), is the Dunning–Kruger effect (Kruger & Dunning, 1999). According to
this effect, individuals with limited competence are more likely to misjudge their performance.
This may indicate that participants in the control group were overconfident, whereas
participants in the experimental group had a more realistic, though still inflated, self-
assessments. Consequently, both groups reported similar levels of perceived success,
despite actual differences in accuracy. This highlights a crucial pedagogical challenge: not
only teaching students how to evaluate GenAI content critically but also helping them
develop a realistic understanding of their capabilities and limitations in doing so. This
metacognitive gap could lead to overconfidence, potentially making students more
vulnerable to sophisticated deepfakes in real-world scenarios. In sum, this study contributes
to the growing body of knowledge on GenAI literacy and critical thinking pedagogy, paving
the way for educational practices that empower students to become discerning consumers
and responsible creators in a GenAI-driven world. Using eye-tracking technology in this
examination contributes to educational technology research by demonstrating how such
measures can reveal the mechanisms through which educational interventions modify
cognitive processing. The use of eye-tracking technology proved to be valuable in revealing
the underlying cognitive processes, offering a deeper understanding of how students engage
with visual information when attempting to detect deepfakes. The integration of eye-tracking
into pedagogical evaluation provides evidence-based foundations for educational
interventions that move beyond declarative knowledge about GenAI into procedural and
strategic knowledge essential for higher education students. Framing the task of visual
discernment as futile risks overlooking a fundamental cognitive step that precedes and
initiates deeper verification processes. While the rapid progression of GenAI might lead
some to argue that the human capacity to distinguish authentic from synthetic media is a
“lost cause,” this perspective neglects the essential role of visual analysis within the broader
framework of critical evaluation. The primary goal of pedagogical interventions, such as the
one examined in this study, is not to create infallible human detectors but to foster this initial
stage of cognitive engagement. By requiring individuals to actively assess the authenticity of
images, such interventions stimulate cognitive and metacognitive processes that are
essential for the development of critical thinking skills. By training students to move from
passive consumption to active scrutiny, we equip them with an essential foundation for
recognizing and questioning misinformation. This study uses eye-tracking to objectively
measure whether such an intervention can shift the underlying cognitive processing
strategies that are essential for navigating the complex GenAI landscape. As GenAI
technologies continue to evolve, higher education institutions must prioritize the
development of the essential 21st-century literacies, integrating cognitive training with
technological solutions to address GenAI-generated misinformation. Educational practice
must balance the development of critical evaluation skills with fostering a realistic awareness
of human limitations, thereby preparing students to become discerning consumers and
responsible creators in an increasingly complex digital information landscape.” (p.13)

Referencia de Articulo 4:
Zhu, W., Li, T., & Li, Q. (2026). The power of beliefs: How generative artificial intelligence
(GenAI) influences academic performance [El poder de las creencias: Cómo la inteligencia
artificial generativa (GenAI) influye en el rendimiento académico]. Acta Psychologica, 236,
106218. [Link]

Conclusiones de Articulo 4:
“Implications. 7.1. Theoretical implications This study constructed an innovative and
comprehensive theoretical model by organically integrating the BAO model and the MISC
model. The proposed framework considers the multidimensional influences of design
expectation fit, ease of use, and design aesthetics on students' beliefs. At the same time, it
systematically examines the complex relationships between belief-driven continuous use
and academic performance, overcoming the limitations of single theoretical models in
explaining how students adopt and utilize GenAI technology. Furthermore, this study
conceptualizes students' beliefs regarding GenAI tools as three distinct dimensions—
hedonic, learning-oriented, and utilitarian. The hedonic dimension captures the enjoyment
and pleasure derived from using GenAI tools; the learning-oriented dimension reflects how
GenAI supports students' learning ability and comprehension; and the utilitarian dimension
emphasizes the perceived usefulness of GenAI in achieving learning goals and improving
academic performance. This categorization provides a nuanced understanding of how
different belief dimensions shape students' acceptance and sustained use of GenAI tools,
thereby addressing gaps in existing research. Finally, this research advances the theoretical
development of technology acceptance in educational contexts. It highlights the critical role
of emotional factors (hedonic) and design aesthetics—dimensions often overlooked in
previous studies that focused primarily on functional aspects. By illustrating how these
factors interact with functional attributes to jointly influence students' decision-making, the
study significantly enriches and extends the explanatory scope of technology acceptance
theory. In doing so, it offers a more comprehensive perspective for understanding student
behavior in AI-driven learning environments and provides a robust foundation for future
research. 7.2. Practical implications The findings of this study provide several practical
implications for optimizing the integration of Generative Artificial Intelligence (GenAI) in
higher education, focusing on design, implementation, and policy recommendations to
enhance its educational effectiveness. First, this study confirms that design expectation fit
significantly influences students' beliefs, highlighting the necessity for GenAI tools to align
with students' learning goals. Educational institutions and developers should refine GenAI's
customization features to provide discipline-specific support. For example, in STEM
education, GenAI should emphasize problem-solving, mathematical modeling, and
computational efficiency, whereas in humanities and social sciences, it should focus on
critical thinking, argumentative structuring, and textual analysis. Additionally, transparency in
content generation processes should be emphasized to build trust and credibility among
students. Second, although ease of use did not directly impact belief formation in this study,
usability remains a fundamental prerequisite for sustained engagement. This finding
suggests that higher education institutions should incorporate targeted training sessions to
enhance students' familiarity with GenAI functionalities. Universities should integrate digital
literacy modules into their curricula to guide students on how to navigate and optimize
GenAI's capabilities in academic tasks. Developers should also prioritize the creation of
user-friendly interfaces with structured guidance mechanisms to minimize the initial cognitive
load associated with GenAI adoption. Third, while design aesthetics did not significantly
shape beliefs in this study, its role in enhancing long-term engagement should not be
overlooked. Developers should ensure that GenAI platforms adopt minimalist, professional,
and distraction-free interfaces that foster a productive learning environment. Instead of
prioritizing purely aesthetic enhancements, focus should be placed on ergonomic interface
design, which supports knowledge organization and cognitive efficiency” (p.9)
“in learning tasks. Fourth, the study highlights the significant role of continuous use in
improving both task performance and technology performance. To facilitate habitual
engagement, higher education institutions should integrate GenAI into Learning
Management Systems (LMS) to ensure seamless accessibility. Faculty members should
actively demonstrate GenAI's utility in coursework, guiding students on its responsible and
effective usage in research, writing, and analytical tasks. Moreover, incorporating peer
learning communities where students share best practices for GenAI usage can enhance
engagement and optimize learning outcomes. Finally, as GenAI adoption expands, ethical
concerns regarding academic integrity and responsible usage must be addressed.
Universities should establish clear AI ethics policies, outlining acceptable and unacceptable
applications of GenAI in academic work. For example, while GenAI can be used as an
assistive tool for idea generation and conceptual clarification, direct AI-generated content
submission should be discouraged. Educators should incorporate AI literacy training,
teaching students how to critically evaluate AI-generated content and ensuring ethical usage
aligned with academic standards. Furthermore, the successful integration of GenAI in higher
education requires collaboration between educators, AI developers, policymakers, and
researchers. Institutions should establish interdisciplinary task forces to continuously refine
AI-enhanced learning strategies based on empirical research. Additionally, policymakers
should develop regulatory frameworks that balance innovation with ethical considerations,
ensuring that GenAI tools align with pedagogical objectives and learning integrity. By
implementing these practical strategies, higher education institutions can maximize the
pedagogical benefits of GenAI while maintaining ethical standards and enhancing students'
learning experiences. This study provides a foundation for further research on the
sustainable and responsible integration of GenAI in academia.” (p.10)

Referencia de Articulo 5:
Teng, D., Zhou, X., Al-Samarraie, H., & Fang, L. (2026). Enhancing critical thinking and self-
efficacy with GenAI: A social cognitive perspective using structural equation modelling
[Mejora del pensamiento crítico y la autoeficacia con IA generativa: Una perspectiva
sociocognitiva mediante modelamiento de ecuaciones estructurales]. Journal of Computer
Assisted Learning, 42(1), 1–18. [Link]

Conclusiones de Articulo 5:
“Discussion This exploratory study examined, using eye-tracking technology, the gaze
patterns of guided exposure to deepfake and authentic portrait images in comparison to
unguided exposure to assess their influence on the acquisition and application of critical
thinking literacy among first-year Bachler students. Eye-tracking represents a central
methodological contribution of this study, as it provides objective indicators of participants’
cognitive processing during the intervention. Prior research has shown that eye-movements
can distinguish learning strategies, support analysis of problem-solving, and capture local
versus global visual attention (Catrysse et al., 2018; Schreiter & Vogel, 2023; Soares et al.,
2021). By applying this method to the context of deepfake detection, the study not only
advances understanding of how guided versus unguided exposure shapes gaze behaviour
but also demonstrates the broader potential of eye-tracking for investigating how educational
interventions influence students’ critical thinking processes. The results, while preliminary,
show a significant difference in visual attention patterns between the experimental and
control groups. These initial findings suggest that targeted training can alter the cognitive
processing strategies students employ when evaluating GenAI-generated content.
Specifically, the experimental group, which received critical thinking intervention, exhibited a
wider distribution of fixations compared to the control group (H1 accepted, H2 partially
accepted). As in previous studies (Cheetham et al., 2013; Gupta et al., 2020; Huang et al.,
2024; Sano et al., 2024; Wohler ¨ et al., 2021), both groups spend similar total fixation
durations across the main facial features of eye, nose and mouth. Focusing on these
features while examining a face is mainly explained by psychological and social aspects,
since these features are critical for conveying emotional expressions, social signals, and
identity recognition (Porras & Louwerse, 2025). However, the finding that the experiment
group significantly expanded their visual attention to peripheral areas (higher total fixation
duration of background AOI’s), may indicate a shift from intuitive visual examination strategy
to a more comprehensive evaluation approach, emphasizing the interventions impact. This
finding is particularly significant given the challenges identified in previous deepfake
detection research (Gupta et al., 2020; Huang et al., 2024). Huang et al. (2024) found that
an image’s background plays a key role in how viewers judge its source. Yet, most visual
attention typically focuses on the central facial features - the eyes, mouth, and nose
(Fearington et al., 2025). Building on this, the current study suggests that the intervention
enhanced students’ critical thinking literacy and encouraged them to broaden their visual
attention to image areas that support more accurate deepfake detection. The significantly
better overall detection accuracy demonstrated by the experiment group in comparison to
the control group (H3 accepted) further supports the intervention’s positive impact on
students’ practical skills in detecting deepfakes. Nevertheless, it is crucial to note that both
groups still exhibited relatively low accuracy for individual images. This outcome aligns with
the literature, which consistently reports the inherent difficulty humans face in distinguishing
between GenAI-generated content and authentic media, especially as GenAI technology
rapidly advances (Huang et al., 2024; Nightingale & Farid, 2022). While the intervention
improved overall accuracy, the persistent challenge at the individual image level
underscores the sophistication of current deepfake algorithms. This suggests that even with
explicit guidance, the visual cues for detection can be subtle, inconsistent, or even absent in
highly convincing deepfakes. The results align with those of Bray et al. (2023), who found
that while an intervention increases accuracy, the overall improvement may not always be
substantial enough to overcome the fundamental difficulty posed by highly realistic synthetic
media. Diel et al. (2024) suggested combining several intervention strategies, such as
feedback training and raising awareness, to improve detection. The implication for education
is that, while training is beneficial, it must be continuous, diverse and adaptive to the
evolving visual capabilities of GenAI. The findings revealed that the control group tended to
classify the images as authentic compared to the experimental group. The control group’s
tendency to classify more images as authentic, even when they were deepfakes, aligns with
the documented AI hyperrealism bias., e.g., a higher tendency to choose GenAI images of
faces as authentic humans (Nightingale & Farid, 2022). Participants” (p.12)
“in the experimental group demonstrated a greater propensity to classify images as
deepfakes compared to the control group (H4 rejected). This result is consistent with findings
by Chen et al. (2025) and Rehman et al. (2025), who also observed that AI detection training
increases participants’ tendency to identify images as fake. These findings carry critical
implications for educational practice. While training can improve visual sampling strategies
and help mitigate biases such as hyperrealism, it may also introduce a new bias toward
over-identifying deepfake. To translate improved information-gathering into more accurate
judgments, additional pedagogical elements are needed. By providing specific principles and
strategies for deepfake detection, the intervention equipped students with cognitive tools to
critically scrutinize images rather than defaulting to an assumption of authenticity. This is a
critical outcome for critical thinking literacy in a GenAI-saturated information environment, as
it demonstrates the potential to mitigate inherent biases that can lead to misinterpretations of
digital content. Contrary to the hypothesis, there was no significant difference between the
experimental and control groups in their perceived success of identification (H5 rejected).
Both groups similarly overestimated their performance, with self-assessed estimates
significantly higher than their actual accuracy. This finding aligns with the existing literature
(Kobis ¨ et al., 2021), which suggests that individuals often struggle to accurately gauge their
performance on complex tasks, especially when guessing is involved (Vuorre & Metcalfe,
2022). This persistent overestimation, even in the experiment group, suggests that while the
intervention improved actual detection skills and altered visual gaze patterns, it did not
necessarily enhance metacognitive awareness of those improvements. It is possible that the
cognitive load associated with the deepfake detection task made it challenging for
participants to accurately monitor their performance. Another explanation, also suggested by
Kobis ¨ et al. (2021), is the Dunning–Kruger effect (Kruger & Dunning, 1999). According to
this effect, individuals with limited competence are more likely to misjudge their performance.
This may indicate that participants in the control group were overconfident, whereas
participants in the experimental group had a more realistic, though still inflated, self-
assessments. Consequently, both groups reported similar levels of perceived success,
despite actual differences in accuracy. This highlights a crucial pedagogical challenge: not
only teaching students how to evaluate GenAI content critically but also helping them
develop a realistic understanding of their capabilities and limitations in doing so. This
metacognitive gap could lead to overconfidence, potentially making students more
vulnerable to sophisticated deepfakes in real-world scenarios. In sum, this study contributes
to the growing body of knowledge on GenAI literacy and critical thinking pedagogy, paving
the way for educational practices that empower students to become discerning consumers
and responsible creators in a GenAI-driven world. Using eye-tracking technology in this
examination contributes to educational technology research by demonstrating how such
measures can reveal the mechanisms through which educational interventions modify
cognitive processing. The use of eye-tracking technology proved to be valuable in revealing
the underlying cognitive processes, offering a deeper understanding of how students engage
with visual information when attempting to detect deepfakes. The integration of eye-tracking
into pedagogical evaluation provides evidence-based foundations for educational
interventions that move beyond declarative knowledge about GenAI into procedural and
strategic knowledge essential for higher education students. Framing the task of visual
discernment as futile risks overlooking a fundamental cognitive step that precedes and
initiates deeper verification processes. While the rapid progression of GenAI might lead
some to argue that the human capacity to distinguish authentic from synthetic media is a
“lost cause,” this perspective neglects the essential role of visual analysis within the broader
framework of critical evaluation. The primary goal of pedagogical interventions, such as the
one examined in this study, is not to create infallible human detectors but to foster this initial
stage of cognitive engagement. By requiring individuals to actively assess the authenticity of
images, such interventions stimulate cognitive and metacognitive processes that are
essential for the development of critical thinking skills. By training students to move from
passive consumption to active scrutiny, we equip them with an essential foundation for
recognizing and questioning misinformation. This study uses eye-tracking to objectively
measure whether such an intervention can shift the underlying cognitive processing
strategies that are essential for navigating the complex GenAI landscape. As GenAI
technologies continue to evolve, higher education institutions must prioritize the
development of the essential 21st-century literacies, integrating cognitive training with
technological solutions to address GenAI-generated misinformation. Educational practice
must balance the development of critical evaluation skills with fostering a realistic awareness
of human limitations, thereby preparing students to become discerning consumers and
responsible creators in an increasingly complex digital information landscape.”(p.13)

Artículo 6
 Acosta-Enriquez, B. G., Ballesteros, M. A. A., Guzman Valle, M. de L. A., Angaspilco,

J. E. M., Aquino Lalupú, J. del R., Jaico, J. L. B., Germán Reyes, N. C., Alarcón

García, R. E., & Castillo, W. E. J. (2025). The mediating role of academic stress,

critical thinking and performance expectations in the influence of academic self-

efficacy on AI dependence: Case study in college students. Computers and

Education: Artificial Intelligence, 8(100381), 100381.

[Link]

Conclusiones:

“This study significantly contributes to the understanding of the mechanisms linking

academic self-efficacy with AI dependency in higher education. The findings reveal complex

interaction patterns between psychological and behavioral factors, moving beyond simplistic

explanations for technology adoption in educational settings. The first significant finding is

that the relationship between academic self-efficacy and AI dependency is mediated by

academic stress, revealing a previously unexplored psychological mechanism in the

literature on educational technology adoption. This discovery suggests that interventions

aimed at reducing technological dependency should address not only cognitive aspects but

also the emotional components of learning. A second innovative finding demonstrates the

serial mediation of academic stress and performance expectations, highlighting a causal

chain that links self-efficacy beliefs with technology use patterns. This nuanced

understanding of the interactions between psychological and behavioral variables provides a

stronger foundation for developing effective educational interventions. The research also

reveals counterintuitive findings regarding the role of critical thinking. The lack of significant

mediating effects from critical thinking challenges prior assumptions about its centrality in

technology adoption and suggests the need to reconsider existing theoretical models on AI

use in higher education. The methodology employed in this study represents an innovation

in educational technology research by integrating serial mediation analysis with structural

equation modeling. This methodological approach provides a replicable template for future

studies on the adoption of emerging technologies in educational contexts. The results

partially refute previous theories that emphasized the predominant role of cognitive

processes in technology adoption, suggesting instead a more complex model where

emotional factors and performance expectations play a crucial role. This refutation opens

new lines of inquiry into the psychological determinants of AI use in education. The

explanatory power of the proposed model (58.9% variance in AI dependency) represents a

significant advancement in quantitatively understanding the factors influencing educational

technology adoption. This level of predictive accuracy provides a solid empirical foundation

for developing institutional policies and intervention programs. The study also contributes to

methodological advancements in the field by validating measurement instruments adapted to

the Latin American context, offering reliable tools for future research in the region. The

cross-cultural validation of these measures represents a significant step forward for

international comparative research. The implications of these findings extend beyond

academia, suggesting the need for a holistic approach in designing educational policies that

consider the interactions among psychological, technological, and pedagogical factors. This

deeper understanding of these mechanisms allows for anticipating and addressing emerging

challenges in AI integration in higher education. In summary, this research not only expands

existing knowledge on AI adoption in higher education but also provides a renewed

conceptual and methodological framework for addressing the challenges of technological

integration in educational contexts. The findings suggest the need to rethink traditional

approaches to technology adoption, considering the complex interplay among psychological

factors, emotional dimensions, and performance expectations” (p.9)

Artículo 7

Contreras Contreras, F., & Olaya Guerrero, J. C. (2025). La inteligencia artificial en la

educación superior peruana: Tendencias y desafíos. Tribunal. Revista en

 Ciencias de la Educación y Ciencias Jurídicas, 5(13).

[Link]

Conclusiones:

“En la presente investigación se evidencia que la integración de inteligencia artificial

(IA) en la educación superior peruana avanza mediante iniciativas fragmentadas,
marcadas por una clara dicotomía entre potencial técnico y barreras estructurales. La
s tendencias destacadas, personalización del aprendizaje, optimización de gestión
universitaria y fortalecimiento investigativo, muestran impactos positivos en eficiencia y
rendimiento académico. No obstante, su alcance se ve limitado por la centralización
geográfica, la infraestructura insuficiente y la escasa articulación interinstitucional, lo cual
profundiza inequidades entre contextos urbanos y rurales.
Adicionalmente, los desafíos críticos trascienden lo tecnológico, radicando en dimensiones
humanas y éticas. La brecha formativa docente, con baja alfabetización digital y resistencia
al cambio, y la ausencia de marcos normativos robustos generan riesgos como
dependencia tecnológica no crítica, vulneración de privacidad y erosión de la originalidad
académica. Estos factores exigen políticas integrales que prioricen capacitación continua
en competencias técnicas y éticas, junto con auditorías algorítmicas que aseguren
transparencia.
Asimismo, la investigación revela una desconexión entre innovación y pertinencia
sociocultural. La baja producción científica local (2% de publicaciones globales) y la omisión
de variables como diversidad lingüística o sostenibilidad financiera reflejan una miopía que
perpetúa modelos importados. Urge fomentar redes colaborativas entre universidades,
Estado e industria para desarrollar soluciones contextualizadas, donde la IA
complemente, no reemplace, el juicio humano y el pensamiento crítico.” (p.57-58)

Artículo 8

Diaz Ancco, F., Rodríguez Gonzales, K., & Estrada Chacón, L. H. (2024). Impacto de

la Inteligencia Artificial en la formación de estudiantes de Educación superior.

Yachay - Revista Científico Cultural, 13(1), 44–61.

[Link]

Conclusiones:

“La IA busca optimizar procesos, automatizar tareas repetitivas y aumentar la

productividad en diversos sectores, incluida la educación, adapta el proceso de
enseñanza-aprendizaje a las necesidades individuales de los estudiantes, ofreciendo
experiencias educativas más personalizadas y [Link] investigación en IA busca
desarrollar nuevas tecnologías y aplicaciones que puedan revolucionar la forma en que
aprendemos, enseñamos y nos relacionamos con el [Link] automatizar
tareas tediosas y repetitivas, liberando tiempo para actividades más creativas y
significativas mejorando la efectividad del proceso de enseñanza aprendizaje y
aumentar las mejores condiciones de retención de los conocimientos. Procesar y analizar
grandes cantidades de datos de manera rápida y eficiente, proporcionando información
valiosa para la toma de decisiones educativas. La IA abre nuevas posibilidades en la
educación, desde la creación de sistemas de tutoría inteligentes hasta la implementación de
aulas virtuales y herramientas de evaluación automatizadas.
Sin embargo, el uso de la IA genera preocupaciones sobre la privacidad y seguridad de los
datos de los estudiantes, así como el riesgo de sesgos algoritmos; dependencia excesiva de
la tecnología, existiendo el riesgo de que la dependencia excesiva de la IA pueda
deshumanizar el proceso educativo y disminuir la importancia de las habilidades
sociales y emocionales.
Ha hecho visible el acceso desigual a la tecnología pudiendo exacerbar las disparidades
educativas, dejando atrás a aquellos que no tienen acceso a recursos tecnológicos
adecuados.
La implicancia y desafío a los que se enfrentan las políticas educativas en el nuevo contexto
de masificación de la IA en el mundo del trabajo académico y la vida social, que está muy
relacionada con el desarrollo de habilidades que le permita a los estudiantes y jóvenes a
desenvolverse apropiadamente en su vida futura ya sea personal como profesional.
En algunos casos se reconoce el desarrollo de capacidades mediante el uso y manejo
de la información general capaz de resolver sus diferentes dudas e inquietudes, para medir
el progreso de sus aprendizajes, plantean soluciones únicas en la toma decisiones con
mayor o menor precisión. Se destaca que la IA optimiza los procesos educativos y
personaliza el aprendizaje de acuerdo con las necesidades individuales de los
estudiantes, enfocado fundamentalmente a su realidad. Finalmente, si bien la Inteligencia
Artificial ofrece numerosas ventajas y oportunidades para mejorar la educación, también
presenta desafíos significativos que deben abordarse de manera cuidadosa y reflexiva
para garantizar su aplicación ética y efectiva en el ámbito educativo” (p.58-59)

ARTÍCULO 9

Fan, L., Deng, K., & Liu, F. (2025). Educational impacts of generative artificial

intelligence on learning and performance of engineering students in China.

Scientific Reports, 15(1), 26521. [Link]

Conclusiones
“This study explores the use of generative AI among engineering students in China,
examining its potential and challenges in engineering education through a questionnaire
survey. The data and findings in this study offer valuable insights that can inform and
support the developments of pedagogical theories. The results indicate that a majority of
students were already using generative AI in their academic and daily learning activities. The
survey shows that generative AI has positively impacted learning efficiency, with 88.52% of
respondents reporting improved productivity through faster content generation and feedback.
64.19% of participants experienced increased learning initiative, though 6.76% felt over-
reliance negatively affected their motivation. Nearly 48% believed AI enhanced independent
thinking, offering new perspectives, while a small minority felt it weakened their critical
thinking. AI also boosted creativity for 58.78% of respondents, though 29.73% saw no
change, and 11.48% felt it stifled creativity due to formulaic content. Nearly half of the
respondents felt that using generative AI did not improve their academic performance,
despite most acknowledging increased learning efficiency and greater engagement in active
learning. Generative AI in education presents several challenges, particularly in engineering,
with students highlighting issues such as the inaccuracy of AI-generated content (62.16%),
over-reliance on AI tools (39.86%), and technical difficulties with usability (20.27%). The
inaccuracy of AI outputs, especially for complex tasks, undermines student confidence, while
concerns about over-dependence may hinder independent problem-solving skills. Ethical
concerns, privacy issues, and high costs were also identified as barriers to AI tool adoption.
Additionally, while 42.57% of students found AI tools suitable for their needs, nearly 40%
expressed doubts about AI’s ability to handle specialized engineering problems, indicating a
need for more adaptable and specialized AI solutions in the field. The survey indicates
strong optimism about the future of generative AI in education, with most respondents
viewing its potential as broad, though 27.7% remained neutral due to challenges and ethical
concerns. Regarding its integration into engineering education, 20.95% supported full
integration, while 43.92% favored partial integration, with many believing generative AI
courses should be compulsory. For regulation, 55.41% favored clear usage guidelines, and
47.3% supported tailored policies based on curriculum needs, with 46.62% advocating for
training for both students and faculty. Opinions on whether AI could replace traditional
teaching methods were mixed, with 41.89% uncertain and 12.84% supporting the idea,
though concerns over technical and ethical issues limited strong support. Respondents also
emphasized the need for AI tools to improve accuracy, literature search capabilities, and
integration with professional software. To move forward, educational institutions should
focus on clear guidelines, customized integration plans, and comprehensive training
programs covering technical, practical, and ethical aspects of generative AI” (p, 12-13).

ARTÍCULO 10

García Gutiérrez, G. F., Tapia Guerrero, J. A., Mejía Peralta, C. E., & Egüez Cevallos,

R. C. (2025). Uso de inteligencia artificial generativa como herramienta de

apoyo académico en la innovación pedagógica universitaria. Revista Social

Fronteriza, 5(6), e9551.

[Link]

Conclusiones
“Esta revisión integradora establece que la inteligencia artificial generativa puede constituir
herramienta efectiva de apoyo académico en innovación pedagógica universitaria cuando su
integración se fundamenta en diseños instruccionales que promuevan pensa
Conocimiento crítico, autonomía estudiantil y acompañamiento docente estratégico,
confirmando la hipótesis central de esta investigación. No obstante, esta efectividad no
emerge automáticamente de disponibilidad tecnológica sino que requiere
transformaciones sistémicas en objetivos formativos, metodologías didácticas, sistemas
de evaluación, políticas institucionales y cultura académica. Las cuatro preguntas de
investigación planteadas encuentran respuestas convergentes: (1) la IAG se integra
principalmente mediante personalización de aprendizaje, apoyo a escritura académica y
tareas de programación, aunque con variabilidad significativa en profundidad y sofisticación
pedagógica; (2) los factores determinantes de efectividad son primordialmente pedagógicos
(nivel de compromiso cognitivo, mediación docente, diseño de tareas) más que
tecnológicos (funcionalidad, eficiencia); (3) los desafíos éticos y prácticos
principales incluyen redefinición de integridad académica, gestión de sesgos
algorítmicos, equidad en acceso y desarrollo de políticas institucionales claras; y (4)
las estrategias de innovación efectivas priorizan rediseño de tareas hacia pensamiento
de orden superior, incorporación curricular explícita de IAG, desarrollo metacognitivo y
formación docente especializada. La literatura analizada demuestra consenso creciente
en qué prohibiciones absolutas resultan inviables y contraproducentes, mientras que la
adopción acrítica sin transformación pedagógica genera riesgos sustanciales,
posicionando la integración reflexiva y diseñada intencionalmente como aproximación
ó[Link] implicaciones de estos hallazgos para práctica educativa, política
institucional e investigación futura son significativas. Instituciones de educación superior
deben transitar de posturas reactivas de prohibición o permisividad ambigua hacia
políticas proactivas que establezcan principios claros de uso responsable, proporcionen
formación docente robusta y rediseñan sistemas de evaluación para privilegiar
competencias difícilmente replicables por IA: pensamiento crítico contextualizado,
creatividad auténtica, juicio ético fundamentado y capacidad de síntesis interdisciplinaria
compleja. Los docentes universitarios requieren apoyo institucional no sólo para
desarrollar competencias técnicas en uso de IAG sino para repensar fundamentalmente
epistemologías disciplinares, identidades profesionales y prácticas pedagógicas en
contextos donde fronteras entre cognición humana y artificial se tornan
progresivamente porosas. La investigación futura debe priorizar estudios
longitudinales que evalúen impactos sostenidos de IAG en desarrollo de
competencias,investigaciones comparativas que analicen efectividad de diferentes diseños
pedagógicos, y estudios contextualizados que examinen cómo variables socioculturales,
institucionales y disciplinares median efectos de estas tecnologías. Particularmente
urgente resulta investigación sobre equidad que documente si IAG amplifica o reduce
brechas educativas preexistentes en contextos de desigualdad estructural. La
inteligencia artificial generativa representa simultáneamente oportunidad transformadora y
riesgo significativo para calidad, equidad e integridad de educación superior: la
trayectoria que se materialice dependerá fundamentalmente de decisiones
pedagógicas, éticas y políticas que comunidades académicas adopten en este
momento crítico de transición tecnológica”.(p, 23-24)

ARTÍCULO 11
 Heredia Pérez, G., Benavides-Gálvez, J. B., Rojas-Campos, E., & Sánchez-Bustamante,

E. (2026). Inteligencia artificial y pedagogía: Retos para la educación superior

en el Perú. Revista InveCom, 6(2). [Link]

Conclusiones
“El estudio confirmó una relación fuerte y significativa entre la inteligencia artificial y la
pedagogía en una universidad privada del Perú, reflejada en un coeficiente de correlación
de 0,859 y una significancia estadística de 0,000. Esto indica que, cuando la inteligencia
artificial se utiliza de manera adecuada, tiene el potencial de fortalecer considerablemente la
pedagogía, lo que a su vez contribuye a optimizar de forma sustancial el proceso de
enseñanza-aprendizaje.
En relación con los principales retos y desafíos en la educación superior, se destacan el
cierre de la brecha digital, la protección de la privacidad y la seguridad, así como el
compromiso ético y la responsabilidad en el uso de estas tecnologías. Para abordar estos
aspectos, es fundamental una gestión cuidadosa que contemple el diseño de políticas
claras y guías prácticas, la colaboración estrecha entre universidades y expertos, y, sobre
todo, la capacitación y el apoyo constante a los docentes para que integren eficazmente la
inteligencia artificial en su práctica pedagógica” (p, 5)

ARTÍCULO 12

Páez, E. F. N., Rodríguez, J. L. G., González, J. A. M., & Gavidia, P. E. G. (2025).

Impact of artificial intelligence on academic performance and its relationship

with dishonesty. Encuentros (Maracaibo).

[Link]

Conclusiones
“La Inteligencia Artificial, si bien representa una herramienta prometedora en la educación,
no debe percibirse como una solución completa a los desafíos actuales en esta área. Su
impacto en el rendimiento académico,según los resultados de este estudio, es relativo y
depende de otros factores como la dedicación del estudiante y la estrategia educativa
aplicada. Los datos reflejan que, aunque las tecnologías de IA pueden facilitar el
aprendizaje y mejorar la motivación, no se traduce directamente en un mejor desempeño
académico, lo que sugiere que otros elementos en el proceso de enseñanza son igualmente
[Link] aprovechar al máximo las ventajas de la IA en el aula, es necesario que los
docentes tengan las habilidades necesarias para aplicar estas herramientas de manera
efectiva,integrándose en su metodología de forma que promuevan un aprendizaje profundo
y significativo. En paralelo, el aprendizaje experiencial y contextual sigue siendo una pieza
clave para una educación completa, que va más allá de la tecnologí[Link] el avance
constante de la IA en el ámbito educativo, resulta fundamental continuar explorando su
evolución y analizar sus efectos en distintos contextos y áreas de estudio. Esto permitirá a
educadores y estudiantes utilizar la IA de forma informada, potenciando sus beneficios y
superando las limitaciones para crear una experiencia educativa más enriquecedora y
adaptada a las necesidades de cada disciplina” (p, 10)

Referencia de Articulo 13:

Ruiz-Rojas, L. I., Salvador-Ullauri, L., & Acosta-Vargas, P. (2024). Collaborative working

and critical thinking: Adoption of generative artificial intelligence tools in higher education
[Trabajo colaborativo y pensamiento crítico: Adopción de herramientas de inteligencia
artificial generativa en la educación superior]. Sustainability, 16 (13), 5367.
[Link]

Conclusiones de Articulo 13:

“The integration of generative AI tools in higher education has captured the attention of
researchers and educators due to its potential to transform traditional pedagogical practices
and promote critical skills. The study found that 64% of the respondents believe that these
tools significantly improve their critical thinking ability. This perception is linked to the tools'
capacity to act as 'collaborative partners' in the learning process, facilitating brainstorming,
problem structuring, and the simulation of complex scenarios that require multi-perspective
analysis” (p.13). “While generative AI tools show promising potential for enhancing
collaboration and critical thinking, the research highlights that their effectiveness is
contingent upon the collaborative framework in which they are deployed. AI does not replace
the need for human interaction; rather, it augments it by handling routine cognitive tasks,
thereby freeing up mental resources for higher-order thinking such as evaluation and
inference. The synergy between 'Collaborative Working' and 'Critical Thinking' is
strengthened when AI is used to mediate group discussions and provide diverse data points
for team analysis” (p.14). “Consequently, the adoption of GenAI should be accompanied by
pedagogical designs that explicitly encourage collaborative problem-solving. Simply
providing access to tools like ChatGPT is insufficient; educators must design tasks that
require students to negotiate meaning, challenge AI-generated outputs collectively, and
synthesize information into coherent arguments. This approach ensures that the technology
serves sustainability in education by fostering resilient, adaptable learners capable of
navigating a digital ecosystem saturated with synthetic media” (p.15).

Referencia de Articulo 14:

Zhang, W., & Liu, X. (2025). Artificial intelligence-generated content empowers college
students’ critical thinking skills: What, how, and why [El contenido generado por inteligencia
artificial empodera las habilidades de pensamiento crítico de los estudiantes universitarios:
Qué, cómo y por qué]. Education Sciences, 15 (8), 977.
[Link]

Conclusiones de Articulo 14:

“Based on the analysis of interview materials and survey questionnaire data, the study
reveals the impacts and mechanisms of college students' use of generative artificial
intelligence tools on their critical thinking skills. The findings of the study are as follows.
First, the frequency of artificial intelligence use is unrelated to critical thinking skills, but
using it for reflective thinking helps to develop critical thinking skills. Simply using the tools
does not guarantee cognitive development; rather, it is the intentional application of these
tools for 'reflective thinking' that correlates with skill enhancement” (p.11). “Second, students
with strong self-regulated learning skills are more likely to use generative artificial
intelligence for reflective thinking and achieve better development in critical thinking skills.
This suggests that self-regulation acts as a catalyst, enabling students to navigate the
affordances of GenAI more effectively. Third, students with strong intrinsic learning
motivation are more likely to use generative artificial intelligence for reflective thinking and
achieve better development in critical thinking skills. Intrinsic motivation drives the depth of
engagement with the technology, moving beyond superficial usage to deep cognitive
processing” (p.12). “The study concludes that GenAI empowers critical thinking not by
replacing human thought, but by serving as a sophisticated interlocutor that challenges
students to verify, analyze, and refine information. The 'What' refers to the access to diverse
viewpoints; the 'How' involves the iterative prompting and validation process; and the 'Why'
connects to the students' drive to leverage technology for epistemic growth. Educational
institutions should therefore focus less on restricting access and more on fostering the self-
regulatory and motivational dispositions that convert GenAI usage into a critical thinking
exercise” (p.13).

Referencia del Artículo 15:

Olaya, A. J. N., & Bajaña, D. I. C. (2025). Transforming critical and creative thinking:
The impact of generative artificial intelligence on higher education Salud, Ciencia y
Tecnología.

[Link]

Conclusión del artículo 15:

“La inteligencia artificial generativa representa una innovación con un impacto

significativo en la educación superior, permitiendo identificar su potencial
transformador en el desarrollo del pensamiento crítico y creativo en los estudiantes”
(p. 8). En este sentido, los autores concluyen que la IAG “facilita la personalización
del aprendizaje, mejora la retroalimentación formativa y fortalece la autonomía del
estudiante”, lo cual contribuye a la construcción de “un entorno educativo más
adaptativo e inclusivo” (p. 8).

De manera complementaria, el estudio señala que “el uso de estas tecnologías

permite una mejor estructuración de los argumentos, la optimización de la
evaluación automatizada y el fortalecimiento de la interacción en los entornos de
aprendizaje” (p. 8). No obstante, se advierte que su implementación “requiere
regulación y supervisión para mitigar los riesgos asociados a los sistemas
algorítmicos, la privacidad de los datos y el uso excesivo de herramientas
automatizadas” (p. 8), reafirmando así “la importancia del papel del docente como
mediador en el proceso educativo” (p. 8).

Asimismo, los autores enfatizan que “la implementación de la IA en educación debe

ir acompañada de estrategias de formación que incentiven el desarrollo del
pensamiento crítico”, evitando que su uso “se limite a la automatización de procesos
cognitivos” (p. 8). En esa línea, se sostiene que “formar a los docentes en el uso de
estas herramientas es fundamental para aprovechar sus beneficios sin comprometer
la calidad de la educación ni la equidad en el acceso a la tecnología” (p. 8).

Finalmente, se concluye que “la integración efectiva de la inteligencia artificial

generativa requiere un enfoque multidisciplinario que combine la tecnología con
principios pedagógicos sólidos”, con el objetivo de “promover un aprendizaje que
estimule la creatividad, el análisis crítico y la autonomía de los estudiantes” (p. 9).
Los autores recomiendan que futuras investigaciones “exploren estrategias
específicas para optimizar su uso y minimizar sus riesgos”, garantizando así “una
educación más equitativa, ética e innovadora” (p. 9).

Referencia del Artículo 16:

Oliveira, L., Tavares, C., Strzelecki, A., & Silva, M. (2025).

Prompting minds: Evaluating how students perceive

generative AI’s critical thinking dispositions. Electronic

Journal of E-Learning, 23(2), 1–18.

[Link]

Conclusión:

“The PCTD-GAI scale fills a significant gap in the literature by offering a validated instrument

for assessing students' perceptions of AI’s critical thinking dispositions. This study provides

novel insights into how students evaluate AI-generated reasoning, systematicity, and truth-

seeking abilities, contributing to ongoing discussions about AI’s role in higher education. As

AI integration in education continues to grow, it is imperative that students develop the ability

to critically assess AI-generated outputs rather than unquestioningly accept them. The

PCTD-GAI scale serves as a foundational tool for educators and researchers to monitor AI

reliance in learning environments and develop pedagogical strategies that foster deeper

critical engagement with AI technologies. Additionally, the scale offers an empirical

foundation for guiding AI literacy programs, ensuring that students are equipped with the
metacognitive skills needed to engage critically with AI tools rather than becoming passive

consumers of AI-generated content.” (Oliveira et al., 2025, p. 12)

Oliveira et al. (2025) “conclude that the PCTD-GAI offers an empirically grounded

mechanism to assess students’ perceptions of generative AI’s critical-thinking dispositions

and to inform pedagogical interventions and AI-literacy programs aimed at reducing

uncritical acceptance of AI outputs. The authors further emphasize methodological

constraints—validation using a single platform (ChatGPT), a sample limited to students from

Portugal and Poland, reliance on self-reported data, and a cross-sectional design—which

warrant broader cross-platform, cross-cultural, mixed-method, and longitudinal research to

confirm and extend the scale’s applicability.” (Oliveira et al., 2025, p. 12)

Referencia de Articulo 17:

Puche-Villalobos, D. J. (2024). La inteligencia artificial y el desarrollo del pensamiento

crítico en estudiantes universitarios. Delectus, 7 (2), 59–67.
[Link]

Conclusiones de Articulo 17:

“La conclusión destaca que la interacción entre la inteligencia artificial y las habilidades
cognitivas superiores se relacionan mejorando la manera de procesar la información. Los
resultados del análisis de correlación de Spearman revelaron una correlación muy
significativa de 0.898, demostrando una fuerte relación positiva entre el uso de herramientas
de IA y el desarrollo del pensamiento crítico. Esto indica que, lejos de atrofiar las
capacidades cognitivas, la IA actúa como un mediador que, bajo las condiciones adecuadas,
potencia la capacidad de análisis y síntesis del estudiante” (p.65). “Sin embargo, es
imperativo reconocer la naturaleza dual de esta tecnología. Si bien facilita el acceso
inmediato a vastos volúmenes de datos, existe el riesgo latente de la superficialidad cognitiva
si el docente no interviene activamente. La tecnología por sí sola no garantiza el aprendizaje
profundo; requiere una estrategia pedagógica que obligue al estudiante a cuestionar la
veracidad y el sesgo de los algoritmos. Por tanto, la IA debe integrarse no como un oráculo
de verdades definitivas, sino como una herramienta de apoyo que requiere supervisión y
validación humana constante” (p.66). “En este sentido, las instituciones educativas deben
reformular sus currículos para incluir competencias digitales avanzadas que vayan más allá
del uso instrumental, enfocándose en la 'alfabetización algorítmica'. El desarrollo del
pensamiento crítico en la era de la IA depende de la capacidad del estudiante para discernir
entre información generada automáticamente y conocimiento validado, una habilidad que se
perfila como la competencia fundamental del siglo XXI” (p.67).

Referencia del Artículo 18:

 Martin, A. J., Collie, R. J., Kennett, R., Liu, D., Ginns, P.,

Sudimantara, L. B., Dewi, E. W., & Rüschenpöhler, L. G.

(2025). Integrating generative AI and load reduction

instruction to individualize and optimize students’ learning.

Learning and Individual Differences, 121(102723), 102723.

[Link]

Conclusión:

“There has been rapid uptake of genAI among users around the world. Uptake in the

education space is often ad hoc or poorly implemented, with the effect of

overwhelming students' cognitive capacity and impeding their learning. Harnessing

cognitive and instructional psychology principles, load reduction instruction (LRI) is a

pedagogical framework that offers guidance for application of genAI that does not

inappropriately burden the learner. We argue that the future of genAIrelated learning

will benefit from attention to LRI principles that can provide significant support for

differentiating learning experiences to better cater for individual differences between

novice and expert learners.” (Martin et al., 2025, p. 7)

Referencia del Artículo 19:

Lee, D., Arnold, M., Srivastava, A., Plastow, K., Strelan, P., Ploeckl,

F., Lekkas, D., & Palmer, E. (2024). The impact of generative

AI on higher education learning and teaching: A study of

educators’ perspectives. Computers and Education: Artificial

Intelligence, 6(100221), 100221.

[Link]
 Conclusión :

“This study, conducted by researchers at a leading university in Australia, surveyed and

interviewed staff from the university’s Education Academy regarding the impact of Artificial

Intelligence (AI) on Higher Education (HE). Data revealed there is no homogenous sentiment

within the university’s staff around AI in HE and a great deal of ambiguity regarding best

practice in response to these recent technological developments. The greatest concerns

among staff were around concepts relating to AI’s potential impact on academic integrity,

however these concerns may be exaggerated. AI has had, in recent months, and will

continue to have a dramatic impact on HE. It is imperative that educators, and management,

stay abreast of developments and continue to commit to ongoing research and discourse

regarding responses to AI, keeping in mind the university’s obligation to serve the students

in preparing them for industry.” (Lee et al., 2024, p. 9)

Referencia del Artículo 20 :

Yeung, R. S. K., Tian, R., Chiu, D. K. W., & Choi, S. P.-M. (2025).

University students’ perceptions on how generative artificial

intelligence shapes learning and research practices: A case

study in Hong Kong. Journal of Academic Librarianship,

51(5), 103082. [Link]

Conclusión :

“Yeung et al. (2025) report that the rapid development of generative AI has significantly

influenced university students’ learning and research activities, with widespread acceptance

of GenAI tools for improving academic productivity and for tasks such as information

retrieval, content summarization, and writing. The authors note, however, concurrent
concerns regarding ethical issues and academic integrity and the potential adverse effects

on development of critical thinking and problem-solving skills. Consequently, they indicate

the need to address risks of over-reliance on GenAI and possible impacts on students’

cognitive and analytical abilities, and they call for further research to determine how GenAI

can be employed effectively and ethically within academic and library contexts.” (Yeung et

al., 2025, p. 9)

Artículo 21

Anders, A. D., & Dux Speltz, E. (2025). Desarrollo de alfabetización en inteligencia

artificial generativa mediante el aprendizaje autorregulado: Un enfoque centrado en
el ser humano. Computers and Education: Artificial Intelligence, 6, 100482.

[Link]

Conclusión del artículo 21:

El estudio de Anders y Dux Speltz (2025) concluye que la alfabetización en

inteligencia artificial generativa en la educación superior no puede limitarse al
dominio técnico de las herramientas, sino que debe integrarse de manera
intencional a los procesos de autorregulación del aprendizaje, ya que “generative AI
literacies require more than technical proficiency; they demand intentional support
for learners’ self-regulated learning processes, including goal setting, monitoring,
and reflection, to ensure that AI tools augment rather than replace human cognition”
(p. 11). Desde un enfoque centrado en el ser humano, los autores destacan que la
integración pedagógica de la IA generativa debe priorizar la agencia del estudiante,
la evaluación crítica de los resultados algorítmicos y la reflexión metacognitiva, dado
que “a human-centered approach to generative AI integration emphasizes learner
agency, critical engagement with AI outputs, and reflective practices that position
students as active decision-makers in their learning processes” (p. 13). Asimismo, el
artículo advierte que, en ausencia de una mediación pedagógica estructurada, el
uso de la IA generativa puede fomentar aprendizajes superficiales y una
dependencia cognitiva, ya que “without structured pedagogical scaffolding,
generative AI risks reinforcing surface-level learning and diminishing opportunities
for deep cognitive engagement” (p. 15). En este sentido, los autores sostienen de
manera indirecta que el desarrollo del pensamiento crítico y de la autonomía
académica depende de estrategias educativas que promuevan la planificación, el
monitoreo y la reflexión consciente sobre el uso de la IA, permitiendo que los
estudiantes analicen, cuestionen y regulen activamente su interacción con estas
tecnologías en contextos académicos (Anders & Dux Speltz, 2025, pp. 12–14).
Artículo 22

Haroud, S., & Saqri, N. (2025). Inteligencia artificial generativa en la educación

superior: Perspectivas de docentes y estudiantes sobre apoyo, sustitución y
alfabetización digital. Education Sciences, 15(4), 396.

[Link]

Conclusión del artículo 22:

El estudio de Haroud y Saqri (2025) evidencia percepciones contrastantes sobre el

uso de la inteligencia artificial generativa en la educación superior, mostrando que
mientras los estudiantes adoptan una postura mayoritariamente positiva, asociando
estas tecnologías con mayores niveles de autonomía, creatividad y accesibilidad, los
docentes expresan reservas significativas respecto a su impacto en las dinámicas
pedagógicas tradicionales y en el desarrollo de habilidades humanas esenciales. En
este sentido, los autores señalan que “students adopt a largely positive stance,
highlighting the advantages in terms of autonomy, creativity and accessibility,
[whereas] teachers express significant reservations, notably about the potential
impact of these technologies on essential human skills and traditional pedagogical
dynamics” (p. 14). Asimismo, los resultados indican que los estudiantes perciben la
IA como un facilitador del rendimiento académico y la innovación, mientras que los
docentes advierten que “over-reliance on AI […] could compromise the development
of general and transferable skills such as critical thinking, collaboration and problem-
solving” (p. 14), lo cual resulta especialmente relevante para el análisis del
pensamiento crítico y la autonomía académica. No obstante, ambos grupos
coinciden en que la IA no puede sustituir el rol fundamental del docente, ya que “AI
cannot replace the fundamental role of teachers, particularly in providing
personalized support and managing learners’ emotional and cognitive needs” (p. 14).
De manera indirecta, el estudio concluye que una integración efectiva de la IA
generativa requiere un enfoque equilibrado, en el que estas herramientas actúen
como complemento y no como reemplazo de las prácticas pedagógicas
tradicionales, destacando además que el fortalecimiento de la alfabetización digital
es una condición indispensable para maximizar sus beneficios y minimizar sus
riesgos en contextos universitarios (Haroud & Saqri, 2025, p. 14). Finalmente, los
autores subrayan que estos hallazgos aportan insumos valiosos para el diseño de
políticas educativas y programas de formación docente orientados a una integración
responsable de la IA, manteniendo los valores humanos como eje central del
proceso educativo y abriendo nuevas líneas de investigación sobre su impacto en el
desarrollo de competencias estudiantiles (p. 14).

Artículo 23

Yusuf, A., Pervin, N., & Román-González, M. (2024). Inteligencia artificial generativa
y el futuro de la educación superior: ¿Amenaza a la integridad académica o reforma
educativa? International Journal of Educational Technology in Higher Education, 21,
21.

[Link]

Conclusión del artículo 23:

El estudio de Yusuf et al. (2024) concluye que existe un alto nivel de conocimiento,
familiaridad y uso previo de herramientas de inteligencia artificial generativa en la
educación superior, destacando que “a significant degree of awareness and
familiarity with GenAI tools” fue evidenciada entre los participantes, quienes además
manifestaron una alta probabilidad de continuar utilizándolas en el futuro,
principalmente para “information retrieval and text paraphrasing” (p. 26). Los autores
señalan que el uso de GenAI presenta beneficios relevantes para los procesos de
aprendizaje universitario, siempre que su implementación sea responsable, ya que
“fostering appropriate and responsible use of GenAI tools can enhance various
aspects of the learning process” (p. 26); sin embargo, advierten que dichos
beneficios coexisten con preocupaciones éticas y académicas que requieren la
formulación de políticas claras y orientadoras. De manera indirecta, el estudio
evidencia que el uso acrítico o desregulado de estas tecnologías puede afectar el
desarrollo de habilidades cognitivas superiores, como el pensamiento crítico y la
toma de decisiones autónomas, especialmente cuando se normaliza la dependencia
de la IA para tareas académicas complejas (Yusuf et al., 2024, pp. 23–25).
Asimismo, los resultados muestran una relación significativa entre la percepción del
uso de GenAI y los contextos culturales, destacando que “our findings indicate a
strong correlation between responses regarding GenAI utilization and cultural
practices”, lo que subraya la necesidad de adaptar las estrategias institucionales a
realidades culturales específicas (p. 26). En este sentido, los autores enfatizan que
las políticas educativas no deben seguir un enfoque homogéneo, ya que “a one-
size-fits-all approach to GenAI integration in HE may not be appropriate”,
proponiendo en su lugar lineamientos flexibles que reconozcan la diversidad cultural
y promuevan una integración ética, equitativa y reflexiva de la inteligencia artificial
generativa, manteniendo el desarrollo del pensamiento crítico y la autonomía
académica como ejes centrales del proceso educativo (p. 26).

Artículo 24

Gao, J., Zhang, J., & Li, Y. (2025). Do AI chatbot-integrated writing tasks influence
writing self-efficacy and critical thinking ability? An exploratory study. Computers and
Education: Artificial Intelligence

[Link]

Conclusión del artículo 24:

Los hallazgos del estudio de Gao, Zhang y Li (2025) permiten concluir que la
integración de tareas de escritura asistidas por chatbots de inteligencia artificial no
genera diferencias estadísticamente significativas en el desarrollo global del
pensamiento crítico ni de la autoeficacia en escritura, aunque sí se observan
mejoras descriptivas en el grupo experimental frente al grupo de control. En este
sentido, los autores señalan de manera directa que “the results revealed no
statistical difference in terms of both critical thinking ability and writing self-efficacy
even with greater increase among the experimental group than that of the control
peers” (Gao et al., 2025, p. 1 lo que evidencia que el uso de chatbots no garantiza
por sí solo un impacto significativo en estas variables psicológicas y cognitivas. No
obstante, un análisis más detallado muestra que el componente de construcción del
lenguaje sí presentó diferencias significativas a favor del grupo experimental, lo cual
sugiere que los estudiantes con menor dominio lingüístico tienden a beneficiarse del
apoyo de la inteligencia artificial para organizar y estructurar sus ideas durante el
proceso de escritura académica. Desde una perspectiva interpretativa, los autores
explican que esta mejora específica se relaciona con el hecho de que los
estudiantes “strive to promote language organization in their writings” cuando
reciben apoyo automatizado en tareas de escritura (Gao et al., 2025, p. 7)Asimismo,
el estudio advierte que la ausencia de efectos significativos en el pensamiento
crítico puede atribuirse a la corta duración de la intervención y a la falta de
experiencia previa de los estudiantes en el uso pedagógico de chatbots, lo que
puede generar una carga cognitiva adicional que dificulte procesos de razonamiento
profundo. En esta línea, los autores sostienen indirectamente que el uso inicial de
herramientas de IA puede aumentar la carga cognitiva extrínseca, limitando el
desarrollo del pensamiento crítico cuando los estudiantes aún no dominan
estrategias metacognitivas para evaluar y filtrar la información generada por la IA
(Gao et al., 2025). En consecuencia, el estudio concluye que la integración efectiva
de chatbots en tareas de escritura debe orientarse hacia un uso formativo y
reflexivo, en el que estas herramientas sean concebidas como recursos para la
transformación cognitiva y no como simples mecanismos de asistencia procedural,
reforzando además el papel mediador del docente en el desarrollo del pensamiento
crítico y la autoeficacia académica.

Referencia de Articulo 25:

Ninghardjanti, P., Umam, M. C., Winarno, A. S., Risanti, N., Jumiyanto, L., & Widodo, W.
(2025). Evaluating the impact of AI on the critical thinking skills among higher education
students by combining the TAM model and critical thinking theory [Evaluando el impacto de
la IA en las habilidades de pensamiento crítico entre estudiantes de educación superior
combinando el modelo TAM y la teoría del pensamiento crítico]. Frontiers in Education, 10,
Article 1719625. [Link]

Conclusiones de Articulo 25:

“The findings emphasize that the TAM (Technology Acceptance Model) constructs
significantly influenced behavioral intention, whereas critical thinking played a crucial
mediating role. Specifically, the perceived usefulness (PU) and perceived ease of use (PEOU)
are not merely drivers of adoption but are intrinsically linked to the students' cognitive
engagement. When students perceive AI tools as useful for their academic goals, they are
more likely to engage in deep processing strategies that enhance critical thinking” (p.12).
“Then, it can be concluded that not only the usability or utility, but also students' strategic
thinking, epistemic vigilance, and intellectual autonomy have a significant impact on the AI
adoption among higher education students. The model demonstrates that critical thinking is
not just an outcome but a prerequisite for effective AI adoption. Students with higher critical
thinking dispositions are better equipped to integrate AI into their learning workflows without
succumbing to passive reliance, using the technology to amplify their analytical capabilities
rather than substitute them” (p.13). “This research contributes a novel perspective to
educational technology literature by validating the 'Meta-TAM' framework. It suggests that
future educational policies should focus on dual-track training: technical proficiency (how to
use the tool) and cognitive proficiency (how to think with the tool). Safeguarding critical
thinking development requires a shift from viewing AI as a content provider to viewing it as a
cognitive scaffold that supports complex reasoning tasks. Thus, the sustainability of AI in
education depends on maintaining the student's role as the primary critical evaluator of
machine-generated outputs” (p.14).