CoChat
Sign Up
Cited by 265Artificial Intelligence in EducationIntelligent Tutoring Systems and Adaptive LearningArtificial Intelligence in Healthcare and Education

Systematic review of artificial intelligence in education: trends, benefits, and challenges

Open access
Garzón, Juan, Patiño, Eddy, Marulanda, Camilo·2025·Multimodal Technologies and Interaction
DOI 10.3390/mti9080084
Open this paper in CoChat — free accountOpen PDF

Abstract

Artificial intelligence (AI) is changing how we teach and learn, generating excitement and concern about its potential to transform education. To contribute to the debate, this systematic literature review examines current research trends (publication year, country of study, publication journal, education level, education field, and AI type), as well as the benefits and challenges of integrating AI into education. This review analyzed 155 peer-reviewed empirical studies published between 2015 and 2025. The review reveals a significant increase in research activity since 2022, reflecting the impact of generative AI tools, such as ChatGPT. Studies highlight a range of benefits, including enhanced learning outcomes, personalized instruction, and increased student motivation. However, there are challenges to overcome, such as students’ ethical use of AI, teachers’ resistance to using AI systems, and the digital dependency these systems can generate. These findings show AI’s potential to enhance education; however, its success depends on careful implementation and collaboration among educators, researchers, and policymakers to ensure meaningful and equitable outcomes.

AI snapshot

Written from the paper’s full text.

TL;DR

This systematic review synthesizes 155 peer-reviewed empirical studies on AI in education, reporting a sharp rise in research since 2022 and especially 2023, alongside benefits such as enhanced learning outcomes, personalized instruction, and motivation. It also identifies persistent challenges including ethical concerns, teacher resistance, and digital dependency, concluding that AI’s success depends on careful, equitable implementation.

Key findings

  • Research activity increased markedly: publication counts rose from 2 studies in 2015 to 35 in 2023 and 54 in 2024, with the paper linking the spike to generative AI tools such as ChatGPT.
  • Most studies focused on higher education (97 studies), followed by secondary education (26), primary education (20), and teachers (12); language learning was the most represented education field (40 studies), followed by ICT (25) and science education (22).
  • Benefits were grouped into 22 types across cognitive, personal, social, and teacher categories; AI was predominantly associated with cognitive and personal gains, especially enhanced learning outcomes, personalized instruction, and increased student motivation.
  • Challenges were grouped into 16 types; the most frequently reported were ethical concerns (23 studies, 14.8%), teacher resistance (19, 12.3%), digital dependence (17, 11.0%), technical difficulties (10, 6.5%), and increased anxiety (9, 5.8%).
  • Studies were conducted in 35 countries, with Asian nations—particularly China and Taiwan—and the United States dominating; Latin America and Africa were relatively absent. The reviewed studies appeared across 37 indexed journals, with *Computers & Education* most prominent.

Methods

  • The review followed PRISMA and Kitchenham and Charters’ three-step process: planning, conducting, and reporting. Searches were run in Web of Science, Scopus, and Taylor & Francis using a defined AI/education/empirical-study search string, completed on 1 July 2025, yielding 3,406 potential studies; backward citation searching added 435.
  • After removing 1,838 duplicates, 2,003 studies were screened by title, abstract, and keywords; 1,432 were excluded, 571 were assessed against inclusion/exclusion criteria, and 155 studies remained for final analysis.
  • Data extraction covered title, authors, year, country, journal, education level, education field, AI type, benefits, and challenges. Thematic analysis used inductive coding in Microsoft Excel, with independent coding by two authors and disagreements resolved by the first author; Cohen’s Kappa was 0.89. Education levels and fields were standardized using ISCED.

Limitations

  • The paper states that grey literature such as conference papers, dissertations, and policy reports was excluded, potentially omitting emerging or context-specific insights; it also states that it does not deeply explore how AI applications align with specific pedagogical models or teaching strategies, nor does it thoroughly examine long-term consequences such as changes in teacher roles, student autonomy, or institutional policies.
  • The paper states that it did not apply a formal framework to assess certainty of evidence; confidence varies by theme, with cognitive and motivational benefits showing moderate to high confidence, while digital overdependence and ethical risks showed lower confidence due to less consistent reporting.
  • Evident from the text: the abstract and methods report 155 studies published between 2015 and 2025, while the conclusion reports 132 empirical studies published between 2015 and 2024, an inconsistency in the reported sample size and date range.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/mti9080084

Cites

62 of 62 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 46 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    The analysis reveals a marked increase in research in 2023.

    “This analysis reveals a marked increase in research in 2023.”
    Quote in the paperSubstantial claim
  • Findingown data

    Only one study included students from postgraduate programs.

    “Only one study included students from postgraduate programs (i.e., Master's or Doctoral programs); therefore, we included it in the higher education category.”
    Quote in the paperNot a substantial claim
  • Findingown data

    Interactive Learning Environments and Education and Information Technologies also stand out as prominent publication venues.

    “Following closely, Interactive Learning Environments and Education and Information Technologies also stand out as prominent publication venues.”
    Quote in the paperNot a substantial claim
  • Findingnone given

    Most studies examined short-term interventions.

    “Most studies examined short-term interventions; however, the growing use of AI in diverse educational contexts suggests the need for further research that explores its long-term integration.”
    Quote in the paperSubstantial claim
  • Findingown data

    Asian nations, particularly China and Taiwan, along with the United States, dominate the field.

    “Asian nations, particularly China and Taiwan, along with the United States, dominate the field.”
    Quote in the paperNot a substantial claim
  • May combine two claims. 30
  • Classified as probably not a substantial claim. 31

  • May combine two claims.

    Findingown data

    The reviewed studies report benefits including enhanced learning outcomes, personalized instruction, and increased student motivation.

    “Studies highlight a range of benefits, including enhanced learning outcomes, personalized instruction, and increased student motivation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The reviewed studies report challenges including students' ethical use of AI, teachers' resistance to using AI systems, and digital dependency.

    “However, there are challenges to overcome, such as students' ethical use of AI, teachers' resistance to using AI systems, and the digital dependency these systems can generate.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The number of studies conducted at the primary, secondary, and teacher education levels was notably lower and relatively evenly distributed.

    “In contrast, the number of studies conducted at the primary, secondary, and teacher education levels was notably lower and relatively evenly distributed.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The studies were conducted in 35 countries from all inhabited continents.

    “The studies were conducted in 35 countries from all inhabited continents.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingcitation

    Studies in ICT often involve teaching students about AI itself or using AI to support algorithmic thinking, coding instruction, or learning analytics.

    “Studies in this area often involve teaching students about AI itself or using AI to support algorithmic thinking, coding instruction, or learning analytics [56].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [56] Huang, 2024, Enhancing Computational Thinking Skills Through Artificial Intelligence Education at a STEAM High School, Sci. Educ. source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Other frequently reported cognitive benefits include improvements in problem-solving, knowledge retention, and critical thinking.

    “Other frequently reported cognitive benefits include improvements in problemsolving, knowledge retention, and critical thinking.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingcitation

    Many educators resist adopting AI due to skepticism, lack of trust, fear of job displacement, or a reluctance to change their teaching methods.

    “Specifically, many educators resist adopting AI due to skepticism, lack of trust, fear of job displacement, or a reluctance to change their teaching methods.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    Ethical concerns and teacher resistance emerged as the most prominent challenges.

    “Ethical concerns and teacher resistance emerged as the most prominent challenges, indicating that the human dimensions of AI adoption, such as privacy, bias, and user acceptance, remain critical barriers.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Embodied and Immersive Systems and Intelligent Tutoring Systems remain less frequent in the reviewed studies.

    “Meanwhile, Embodied and Immersive Systems and Intelligent Tutoring Systems remain less frequent, despite strong evidence of their effectiveness in fostering engagement and deep learning [32].”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    There is a strong dominance of language learning, followed by ICT and science education.

    “There is a strong dominance of language learning, followed by ICT and science education.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    Cognitive and motivational benefits of AI were frequently reported across multiple contexts and supported by converging evidence, suggesting moderate to high confidence in these themes.

    “For example, cognitive and motivational benefits of AI were frequently reported across multiple contexts and supported by converging evidence, suggesting moderate to high confidence in these themes.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Generative AI was the most implemented AI type in the reviewed studies, followed by Conversational and NLP Agents, and Machine Learning Models.

    “The analysis shows that Generative AI was the most implemented type, followed by Conversational and NLP Agents, and Machine Learning Models.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    The results show a multifaceted range of challenges associated with integrating AI into educational environments, with personal and teacher-related concerns being the most frequently reported.

    “The results show a multifaceted range of challenges associated with integrating AI into educational environments, with personal and teacher-related concerns being the most frequently reported.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    All included studies were peer-reviewed.

    “While all included studies were peer-reviewed”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    AI-related educational research increased significantly, especially from 2023 onward, coinciding with the widespread release of tools such as ChatGPT.

    “The findings reveal a significant increase in AI-related educational research, especially from 2023 onward, coinciding with the widespread release of tools such as ChatGPT.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Occasional discrepancies in screening were resolved through consensus.

    “Occasional discrepancies in this process were resolved through consensus.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Before synthesizing the findings, all extracted data were carefully reviewed to ensure they were complete and consistent.

    “Before synthesizing the findings, we carefully reviewed all extracted data to ensure they were complete and consistent.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    A standardized data extraction form was developed by the research team to ensure consistency in coding and interpretation.

    “A standardized data extraction form was developed by the research team to ensure consistency in coding and interpretation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Themes were identified through iterative comparison, and disagreements were resolved in consultation with the first author.

    “The themes were identified through iterative comparison, and disagreements were resolved in consultation with the first author.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    The predominance of higher education in AI-related research aligns with prior reviews, which attribute this trend to the increased availability of digital infrastructure, funding mechanisms, and institutional autonomy in universities.

    “The predominance of higher education in AI-related research aligns with prior reviews in the field of educational technology, which attribute this trend to the increased availability of digital infrastructure, funding mechanisms, and institutional autonomy in universities [36].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [36] Guo, 2024, Artificial Intelligence in Education Research during 2013–2023: A Review Based on Bibliometric Analysis, Educ. Inf. Technol. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    The review protocol was developed internally by the research team during the planning stage and registered on the Open Science Framework (OSF) platform.

    “This protocol was developed internally by the research team during the planning stage and registered on the Open Science Framework (OSF) platform.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The concentration of publications in these high-impact journals highlights the growing academic recognition of AI's transformative potential in education.

    “The concentration of publications in these high-impact journals highlights the growing academic recognition of AI's transformative potential in education.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    The data extraction form was not formally piloted but was collaboratively refined during the initial phase of data extraction.

    “Although the form was not formally piloted, it was collaboratively refined during the initial phase of data extraction.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The second and third authors independently coded the extracted qualitative data manually using Microsoft Excel.

    “The second and third authors independently coded the extracted qualitative data manually using Microsoft Excel.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    The diversity of journals suggests that the topic transcends traditional disciplinary boundaries, attracting attention from both educational technology specialists and researchers in broader pedagogical and computational domains.

    “This diversity suggests that the topic transcends traditional disciplinary boundaries, attracting attention from both educational technology specialists and researchers in broader pedagogical and computational domains [37].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [37] Wang, 2024, Artificial Intelligence in Education: A Systematic Literature Review, Expert Syst. Appl. source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    The uneven geographic representation raises concerns about whether AI tools can effectively meet the diverse needs of educational systems worldwide.

    “The uneven geographic representation raises concerns about whether AI tools can effectively meet the diverse needs of educational systems worldwide.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationcitation

    Without including diverse educational contexts, it is difficult to ensure that AI solutions are adaptable, ethical, and relevant on a global scale.

    “Without including diverse educational contexts, it is difficult to ensure that AI solutions are adaptable, ethical, and relevant on a global scale [46,47].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [46] Holmes, W., Bialik, M., and Fadel, C. (2023). Artificial Intelligence in Education. Data Ethics: Building Trust: How Digital Technologies can Serve Humanity, Globethics Publications. source

      Not checked in the free run.

    • cites [47] Yambal, S., and Waykar, Y.A. (2025). Future of Education Using Adaptive AI, Intelligent Systems, and Ethical Challenges, IGI Global Scientific Publishing. source

      Not checked in the free run.

    • Rests on an assumption: Educational systems differ widely in terms of language, infrastructure, curricula, and student needs.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    This systematic review adopted the PRISMA statement approach and followed the recommendations of Kitchenham and Charters.

    “This systematic review adopted the PRISMA statement approach [34] and followed the recommendations of Kitchenham and Charters [35].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [34] Page, 2021, The PRISMA 2020 Statement: An Updated Guideline for Reporting Systematic Reviews, Syst. Rev. source

      Not checked in the free run.

    • cites [35] Kitchenham, B., and Charters, S. (2007). Guidelines for Performing Systematic Literature Reviews in Software Engineering Version 2.3.

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    In cases where descriptive details such as country of origin were not explicitly provided, the first author's affiliation was used as a proxy.

    “In cases where certain descriptive details, such as the country of origin, were not explicitly provided, we used the first author's affiliation as a proxy.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Premisenone given

    Most existing reviews of AI in education are limited to pre-ChatGPT publications or narrow scopes focused on specific applications, user groups, pedagogical models, or types of AI systems.

    “Although several reviews have explored AI's role in education, most are limited to pre-ChatGPT publications or narrow scopes focused on specific applications, user groups, pedagogical models, or types of AI systems.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    If the second and third authors disagreed on any specific paper, the first author reviewed the paper and made the final decision to maintain consistency and rigor.

    “If the second and third authors disagreed on any specific paper, the first author reviewed the paper and made the final decision to maintain consistency and rigor.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Interpretationcitation

    There is a risk that AI may be applied inconsistently or inequitably as its adoption increases without guidance.

    “Without such guidance, there is a risk that AI may be applied inconsistently or inequitably as its adoption increases [43,44].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [43] Costa, C.J., Aparicio, M., Aparicio, S., and Aparicio, J.T. (2024). The Democratization of Artificial Intelligence: Theoretical Framework. Appl. Sci., 14. source

      Not checked in the free run.

    • cites [44] Zou, 2025, Navigating the Future: Establishing a Framework for Educators’ Pedagogic Artificial Intelligence Competence, Eur. J. Educ. source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    Education levels and subject areas were standardized using the International Standard Classification of Education (ISCED) to support meaningful comparisons across studies.

    “To support meaningful comparisons across studies, we standardized education levels and subject areas using the International Standard Classification of Education (ISCED).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The review does not thoroughly examine the long-term consequences of AI integration, such as changes in teacher roles, student autonomy, or institutional policies.

    “Third, although the results raised ethical concerns and digital dependency, we do not thoroughly examine the long-term consequences of AI integration, such as changes in teacher roles, student autonomy, or institutional policies.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Previous studies by Strielkowski et al. and Xu emphasize the transformative potential of AI in improving academic outcomes through adaptive learning environments.

    “This aligns with previous studies, such as those by Strielkowski et al. [11] and Xu [57], which emphasize the transformative potential of AI in improving academic outcomes through adaptive learning environments.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [11] Strielkowski, 2024, AI-driven Adaptive Learning for Sustainable Educational Transformation, Sustain. Dev. source

      Not checked in the free run.

    • cites [57] Xu, 2024, AI in Education: Enhancing Learning Experiences and Student Outcomes, Appl. Comput. Eng. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    This review did not apply a formal framework to assess the certainty of the evidence for each outcome.

    “Furthermore, this review did not apply a formal framework to assess the certainty of the evidence for each outcome.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationreasoning

    These tools offer solutions to language learning challenges, such as vocabulary acquisition, grammar correction, and pronunciation feedback, which may explain their popularity among researchers.

    “These tools offer solutions to language learning challenges, such as vocabulary acquisition, grammar correction, and pronunciation feedback, which may explain their popularity among researchers.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    The findings may slightly overrepresent more established academic perspectives and underrepresent practitioner-driven innovations.

    “As such, the findings may slightly overrepresent more established academic perspectives and underrepresent practitioner-driven innovations.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

    Interpretationreasoning

    This trend is not surprising given that many AI applications originate within ICT-related fields and are often conducted by technically proficient educators or within computing departments.

    “This trend is not surprising given that many AI applications originate within ICT-related fields and are often conducted by technically proficient educators or within computing departments.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Many AI applications originate within ICT-related fields and are often conducted by technically proficient educators or within computing departments.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The review does not deeply explore how different AI applications align with specific pedagogical models or teaching strategies across disciplines and contexts.

    “Second, while the review identifies general benefits and challenges of AI in education, it does not deeply explore how different AI applications align with specific pedagogical models or teaching strategies across disciplines and contexts.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Implementing AI in primary and secondary contexts often requires adaptations to pedagogy and content, which may dissuade rapid experimentation.

    “Additionally, implementing AI in these contexts often requires adaptations to pedagogy and content, which may dissuade rapid experimentation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Previous studies emphasize that trust, transparency, and ethical governance are foundational for the successful use of AI in education.

    “These findings are consistent with previous studies [64,65], which emphasize that trust, transparency, and ethical governance are foundational for the successful use of AI in education.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [64] Shukla, 2024, Principles Governing Ethical Development and Deployment of AI, Int. J. Eng. Bus. Manag. source

      Not checked in the free run.

    • cites [65] Fu, 2024, Navigating the Ethical Terrain of AI in Education: A Systematic Review on Framing Responsible Human-Centered AI Practices, Comput. Educ. Artif. Intell. source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    These dualities highlight the importance of considering contextual, technological, and individual variables when evaluating AI's impact.

    “These dualities, where certain factors are reported as strengths and weaknesses, highlight the importance of considering contextual, technological, and individual variables when evaluating AI's impact.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    Compared to earlier studies, this analysis offers a deeper categorization and highlights the tension between AI's potential and its limitations.

    “Compared to earlier studies [19,37], this analysis offers a deeper categorization and highlights the tension between AI's potential and its limitations.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [19] Dimitriadou, 2023, A Critical Evaluation, Challenges, and Future Perspectives of Using Artificial Intelligence and Emerging Technologies in Smart Classrooms, Smart Learn. Environ. source

      Not checked in the free run.

    • cites [37] Wang, 2024, Artificial Intelligence in Education: A Systematic Literature Review, Expert Syst. Appl. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Research by Younas et al. and Wang et al. found that intelligent tutoring systems and gamified AI applications promote student interest and self-directed learning.

    “These findings align with research by Younas et al. [59] and Wang et al. [37], which have also found that intelligent tutoring systems and gamified AI applications promote student interest and self-directed learning.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [59] Younas, 2025, A Comprehensive Systematic Review of AI-Driven Approaches to Self-Directed Learning, IEEE Access source

      Not checked in the free run.

    • cites [37] Wang, 2024, Artificial Intelligence in Education: A Systematic Literature Review, Expert Syst. Appl. source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationcitation

    Our classification of AI benefits into cognitive, personal, social, and teacher domains offers a more comprehensive framework than earlier reviews, which often focused narrowly on learning outcomes.

    “Our classification of AI benefits into cognitive, personal, social, and teacher domains offers a more comprehensive framework than earlier reviews [16][17][18][19][20][21], which often focused narrowly on learning outcomes.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [16][17][18][19][20][21]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationreasoning

    These gaps point to potential missed opportunities for innovation.

    “These gaps point to potential missed opportunities for innovation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    This gap underlines a critical area for future research.

    “Given that teachers play a central role in mediating the integration of AI into classrooms, this gap underlines a critical area for future research.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Teachers play a central role in mediating the integration of AI into classrooms.

  • May combine two claims.

    Interpretationreasoning

    Digital dependence and increased anxiety point to potential unintended consequences of AI overuse, such as reduced creativity and heightened psychological stress, particularly when AI systems are poorly designed or overly focused on assessment.

    “From a cognitive perspective, digital dependence and increased anxiety point to potential unintended consequences of AI overuse, such as reduced creativity and heightened psychological stress, particularly when AI systems are poorly designed or overly focused on assessment.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Researchers such as Strielkowski et al. caution against overly technocentric approaches that neglect pedagogical and emotional considerations.

    “These challenges have been highlighted by researchers such as Strielkowski et al. [11], who caution against overly technocentric approaches that neglect pedagogical and emotional considerations.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [11] Strielkowski, 2024, AI-driven Adaptive Learning for Sustainable Educational Transformation, Sustain. Dev. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    AI is widely perceived as a powerful tool for personalizing instruction, adapting to individual learning needs, and fostering student engagement.

    “These findings suggest that AI is widely perceived as a powerful tool for personalizing instruction, adapting to individual learning needs, and fostering student engagement.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Studies demonstrate that AI can streamline administrative duties, generate personalized learning analytics, and support instructional planning.

    “These results align with studies that demonstrate how AI can streamline administrative duties, generate personalized learning analytics, and support instructional planning [15,17].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [15] Holmes, 2022, State of the Art and Practice in AI in Education, Eur. J. Educ. source

      Not checked in the free run.

    • cites [17] Zhai, 2021, A Review of Artificial Intelligence (AI) in Education from 2010 to 2020, Complexity source

      Not checked in the free run.

  • May combine two claims.

    Interpretationnone given

    Benefits like digital literacy and accessibility demonstrate AI's role in equipping students with essential 21st-century skills and making learning more inclusive for diverse populations.

    “Additionally, benefits like digital literacy and accessibility demonstrate AI's role in equipping students with essential 21st-century skills and making learning more inclusive for diverse populations.”
    Quote in the paperNot a substantial claim
53 of 53

  • The paper relies on this without support.

    Teachers play a central role in mediating the integration of AI into classrooms.

    “Given that teachers play a central role in mediating the integration of AI into classrooms, this gap underlines a critical area for future research.”

  • The paper relies on this without support.

    The release of ChatGPT and subsequent generative AI systems marked a milestone in education, ushering in the post-ChatGPT era.

    “The release of ChatGPT and subsequent similar generative AI systems, such as DeepSeek, Claude, Gemini, and Copilot, has marked a milestone in education, ushering in what might be called the post-ChatGPT era.”

  • The paper relies on this without support.

    Educational systems differ widely in terms of language, infrastructure, curricula, and student needs.

    “Educational systems differ widely in terms of language, infrastructure, curricula, and student needs.”

  • The paper relies on this without support.

    Many AI applications originate within ICT-related fields and are often conducted by technically proficient educators or within computing departments.

    “This trend is not surprising given that many AI applications originate within ICT-related fields and are often conducted by technically proficient educators or within computing departments.”

  • The paper relies on this without support.

    Educators play a central role in mediating the implementation of AI tools.

    “This is notable given the central role of educators in mediating the implementation of AI tools.”

  • The paper relies on this without support.

    Teachers are essential mediators of AI adoption.

    “Given that teachers are essential mediators of AI adoption, more work is needed to empower educators with tools that enhance, not replace, their professional agency and pedagogical expertise.”
6 of 6

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The central boundary condition is that learning benefits depend on how learners interact with mediation-when students engage with AI outputs by questioning, comparing drafts, rejecting or adapting suggestions, and justifying compositional choices, the interaction constitutes active mediated learning rather than passive text transfer.

    “However, the central boundary condition is that learning benefits depend on how learners interact with mediation—when students engage with AI outputs by questioning, comparing drafts, rejecting or adapting suggestions, and justifying compositional choices, the interaction constitutes active mediated learning rather than passive text transfer (Garzón et al., 2025; Ilgun Dibek et al., 2025).”

    From From tool to scaffold: structured human–AI collaboration and its effects on academic writing and digital critical thinking among Saudi EFL learners (2026) · cites this paper as Garzón et al., 2025

    Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

    Experimental — accuracy not yet measured

  • Systematic reviews of AI in education describe accelerating use of AI tools for adaptive learning, intelligent tutoring, automated assessment feedback, and writing support, with reported associations to enhanced learner engagement, higher feedback quality, and a broad range of cognitive outcomes.

    “Systematic reviews of AI in education describe accelerating use of AI tools for adaptive learning, intelligent tutoring, automated assessment feedback, and writing support, with reported associations to enhanced learner engagement, higher feedback quality, and a broad range of cognitive outcomes (Garzón et al., 2025; Ilgun Dibek et al., 2025).”

    From From tool to scaffold: structured human–AI collaboration and its effects on academic writing and digital critical thinking among Saudi EFL learners (2026) · cites this paper as Garzón et al., 2025

    Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Conclusion reports a different review size and date range

    Minor

    The conclusion says the review examined 132 studies published through 2024, whereas the abstract and methods report 155 included studies published through 2025. The 132 figure matches the sum of the annual counts shown for 2015-2024, which explicitly exclude 2025 studies, but it is not the size of the full review.

  • Five dimensions claimed, but six listed

    Minor

    The sentence says five dimensions were examined but names six: year, country, journal, educational level, education field, and AI type. The research-question description also lists all six.

2 of 2

Cites

Flagged first, then by citations

3 of the 65 references couldn’t be matched to a known paper.

Research works printed but not linked: 1

  1. Cole, M., John, V.-S., Scribner, S., and Souberman, E. (1978). Mind in Society: The Development of Higher Psychological Processes, Harvard University Press.

Cited by

Sorted by citations

Related

Authors
Garzón, J., Patiño, E., & Marulanda, C.
Affiliations
Universidad Católica de Oriente (CO)
Corresponding author
Garzón, Juan
ORCID
Venue
Multimodal Technologies and Interaction
Publisher
MDPI AG
Published
2025-08-20
Volume
9
Issue
8
Pages
84
ISSN
2414-4088 (ISSN-L)
DOI
10.3390/mti9080084
OpenAlex
W4413401681
Semantic Scholar
594051870a74e72b85f42c7f52be0954184c81ca
Cited by
265
References
65
Citations by year
2026: 238 · 2025: 27
FWCI
40.73
Type
journal article
Language
en
Open access
Licence
Locations
  • Multimodal Technologies and Interaction (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • OSF Preprints (OSF Preprints) (submittedVersion)
Topics
  • Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
  • Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
artificial intelligence in education, generative AI, ChatGPT, personalized instruction, student motivation, learning outcomes, ethical use of AI, digital dependence, research trends, educational equity

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review