A call for caution and evidence–based research on the impact of artificial intelligence in education

Abstract

Purpose

This paper aims to examine the complex balance between enthusiasm and skepticism regarding artificial intelligence (AI) integration in educational practices. It advocates for a cautious, evidence-based approach while addressing both opportunities and challenges, aligning with the United Nations Sustainable Development Goal 4 (SDG4) for Quality Education.

Design/methodology/approach

Through critical analysis of current discourse surrounding AI in education, this paper synthesizes existing literature on both supportive and skeptical perspectives. The methodology involves systematic examination of past educational technology trends, current AI developments and their implications for teaching and learning. The paper develops its research agenda through careful consideration of existing empirical studies, theoretical frameworks and identifying gaps in current understanding.

Findings

The analysis reveals that while AI offers promising potential for enhancing learning outcomes and educational accessibility, its integration presents significant challenges that require careful consideration. The paper identifies critical tensions between technological innovation and pedagogical values, highlighting areas where enthusiasm for AI adoption must be tempered with empirical evidence and critical evaluation. Current evidence suggests that successful AI integration requires balanced consideration of both opportunities and limitations, with particular attention to maintaining human-centered educational practices.

Originality/value

This viewpoint provides a comprehensive framework for understanding the dialectic between AI’s educational potential and its limitations. By synthesizing both supportive and critical perspectives, it offers a nuanced approach to AI integration that acknowledges both opportunities and challenges. The article’s value lies in its systematic identification of key research priorities and its emphasis on evidence-based implementation strategies that serve educational goals while mitigating potential risks.

Citation

Sposato, M. (2025), "A call for caution and evidence–based research on the impact of artificial intelligence in education", Quality Education for All, Vol. 2 No. 1, pp. 158-170. https://doi.org/10.1108/QEA-09-2024-0087

Introduction

The rapid advancement and proliferation of artificial intelligence (AI) technologies have sparked significant debate within educational sectors, generating both enthusiasm and concern among practitioners, researchers and policymakers. This viewpoint article, directed primarily at educational researchers and policy makers, addresses the pressing need for a balanced, evidence-based approach to understanding and implementing AI in educational contexts. While recognizing the potential benefits of AI integration, this paper critically examines both opportunities and challenges, drawing on extensive literature from both proponents and skeptics of educational technology innovation.

The allure of AI in education stems from its promised capabilities: personalized learning experiences, automated assessment and enhanced accessibility. Early adopters have indeed reported improvements in adaptive learning systems, intelligent tutoring and educational data analytics. However, these initial successes must be contextualized within the broader history of educational technology implementation. Previous technological innovations, from educational television to massive open online course (MOOCs), have demonstrated that initial enthusiasm often outpaces empirical evidence of long-term effectiveness (Gibson and Tesone, 2001; Davies, 1999).

The integration of AI in education represents more than a mere technological advancement; it constitutes a fundamental shift in pedagogical approaches and educational relationships. While AI technologies can enhance metacognition and technological skills, offering new avenues for student engagement and self-directed learning (ElSayary, 2024a), critical voices in the field raise valid concerns about potential negative impacts on critical thinking, creativity and human interaction in education (Biesta, 2010). These competing perspectives reflect a broader tension in educational technology adoption between innovation and preservation of core educational values.

Current research reveals a complex landscape where AI’s educational impact varies significantly across different contexts and applications. For instance, studies by Rahiman and Kodikal (2024) demonstrate AI’s potential to revolutionize higher education through personalized learning pathways. However, their work also highlights crucial challenges in implementation, including issues of equity, access and pedagogical effectiveness. Similarly, Brown et al. (2024) explore how AI might influence idea formation and critical thinking, raising important questions about the technology’s role in shaping student cognition and learning processes.

This paper seeks to bridge the gap between enthusiastic adoption and skeptical resistance by advocating for a research-based approach to AI integration in education. Drawing from both empirical studies and theoretical frameworks, it identifies key areas requiring further investigation while acknowledging the limitations and potential risks of AI implementation. The research directions proposed herein emerge from systematic analysis of existing literature, identified gaps in current understanding and critical evaluation of emerging challenges in educational AI applications.

The need for caution

The imperative for cautious implementation of AI in education emerges from both historical precedent and contemporary research findings. While acknowledging AI’s transformative potential, a critical examination of past educational technology initiatives reveals patterns that warrant careful consideration. Gibson and Tesone's (2001) seminal work on management fads in education provides a theoretical framework for understanding how enthusiastic adoption of new technologies can sometimes overshadow critical evaluation of their educational value. Their research demonstrates how initial excitement about technological innovations often leads to widespread adoption before sufficient evidence of effectiveness has been established.

Education’s inherent complexity distinguishes it from purely technical domains, necessitating particular care in AI implementation. Biesta's (2010) foundational work on evidence-based education emphasizes the intricate interplay between human factors, cognitive processes and socio-emotional development in learning environments. Recent studies by Wu and Yu (2024) build upon this foundation, presenting meta-analytical evidence that while AI chatbots can enhance certain aspects of learning, their effectiveness varies significantly across different educational contexts and student populations. Their findings underscore the need for nuanced understanding of when and how AI tools can most effectively support educational objectives.

Historical analysis of educational technology adoption provides valuable insights for current AI implementation efforts. The trajectory of educational television in the 1950s, as documented by Davies (1999), offers particularly relevant parallels. Initial claims of revolutionary impact gave way to more modest assessments as research revealed the technology’s limitations and the crucial role of pedagogical integration. Similarly, Mitchell and Sutherland's (2020) examination of MOOCs demonstrates how initial enthusiasm for technological solutions often overlooks crucial aspects of effective teaching and learning, particularly in special and inclusive education contexts.

Contemporary research by Williamson (2024) on the social life of AI in education reveals complex interactions between technological capabilities and educational needs. Their work demonstrates how AI systems, while powerful, can inadvertently reinforce existing educational inequalities if not implemented with careful attention to diverse student needs and learning contexts. This finding aligns with Rahiman and Kodikal's (2024) research, which emphasizes the need for thoughtful integration of AI tools to enhance rather than replace human-centered educational practices.

Emerging evidence from Rappleye et al. (2024) suggests that the rapid pace of AI development raises specific concerns about unintended consequences for critical thinking skills, creativity and human elements of education. Their work provides empirical support for more measured approaches to AI adoption, particularly in contexts where traditional educational values and practices may be at risk. These findings are further supported by Kim's (2024) research on teacher–AI collaboration, which identifies potential challenges in maintaining meaningful human interaction within AI-enhanced learning environments.

The integration of AI technologies in education must be viewed within the broader context of educational goals and values. Lockman and Schirmer's (2020) comprehensive review of online instruction in higher education provides a framework for evaluating AI implementations against established principles of effective teaching and learning. Their research emphasizes the importance of maintaining pedagogical integrity while leveraging technological innovations, a balance that becomes increasingly crucial as AI capabilities expand.

Alignment with SDG4: quality education

The integration of AI in education must be evaluated against the framework provided by the United Nations Sustainable Development Goal 4 (SDG4), which aims to “ensure inclusive and equitable quality education and promote lifelong learning opportunities for all” (United Nations, 2015). This alignment requires careful consideration of both opportunities and potential obstacles, particularly in light of emerging research on educational technology implementation and equity concerns.

Recent work by Krupar and Taneja (2020) provides critical insights into the challenges of achieving SDG4 targets through technological innovation. Their research demonstrates that while AI has the potential to enhance educational accessibility and personalization, implementation without careful consideration of existing inequalities may exacerbate rather than ameliorate educational disparities. This finding is particularly relevant when considered alongside Williamson's (2024) analysis of AI’s social impact in educational settings, which reveals complex interactions between technological innovation and existing socioeconomic barriers.

The potential of AI to democratize access to high-quality educational resources requires careful examination. Brown et al. (2024) present evidence that AI-powered systems can break down geographical and socioeconomic barriers through features such as multilingual content delivery and adaptive learning pathways. However, their research also highlights significant challenges in ensuring equitable access to these technologies. The digital divide, as documented by Lockman and Schirmer (2020), remains a substantial barrier to achieving SDG4’s goals of inclusive education through technological means.

AI-powered adaptive learning systems offer promising opportunities for personalized, continuous learning experiences. ElSayary's (2024b) research on leveraging ChatGPT for transformative learning demonstrates how AI can support individualized learning paths and enhance engagement. However, this potential must be balanced against privacy concerns and algorithmic bias, as highlighted by Wu and Yu's (2024) meta-analysis of AI chatbot effectiveness in educational settings. Their work reveals significant variations in outcomes across different student populations, emphasizing the need for careful attention to equity in AI system design and implementation.

The relationship between AI integration and educational quality requires particular attention. Research by Owusu-Agyeman (2024) on lifelong learning in higher education institutions demonstrates how AI can enhance educational quality through personalized feedback and adaptive content delivery. However, their work also emphasizes the critical importance of maintaining human elements in education, particularly in developing social-emotional skills and critical thinking abilities. This aligns with Kim's (2024) findings on teacher–AI collaboration, which suggest that optimal educational outcomes are achieved when AI augments rather than replaces human teaching.

The development of new teaching competencies emerges as a crucial factor in aligning AI implementation with SDG4 goals. Sposato's (2024) research on leadership training in the AI era identifies specific competencies required for effective AI integration in educational settings. Their work, combined with Terry et al.'s (2024) model for teaching evidence-based practice, provides a framework for developing teacher training programs that balance technological literacy with pedagogical expertise.

Pietsch et al.'s (2024) research on open innovation in schools offers valuable insights into how educational institutions can approach AI integration while maintaining focus on SDG4 objectives. Their work demonstrates the importance of collaborative approaches to innovation, involving diverse stakeholders in the development and implementation of AI-enhanced educational practices. This finding is particularly relevant when considered alongside Rappleye et al.'s (2024) analysis of evidence-based SDG4 discussions, which emphasizes the need for inclusive dialogue in educational technology implementation.

The potential for AI to support immersive learning experiences about global sustainability challenges represents another important alignment with SDG4 objectives. However, as Mitchell and Sutherland (2020) demonstrate in their work on special and inclusive education, the effectiveness of such technologies in promoting deep understanding requires careful evaluation and evidence-based implementation strategies. This caution is reinforced by Brown et al.'s (2024) research on preventing idea “echo-chambers,” which highlights the importance of ensuring AI-enhanced learning environments promote diverse perspectives and critical engagement.

Key areas for research

The integration of AI in education demands a rigorous research agenda grounded in evidence-based practice, as established by Rousseau (2006) and further developed by Barends and Rousseau (2018) in their work on evidence-based management. Through systematic analysis of existing literature and identified gaps in current understanding, several critical research priorities emerge that require sustained scholarly attention.

Longitudinal studies of learning impact

The need for extended temporal analysis of AI’s educational impact emerges as a primary research priority. Current research, such as ElSayary's (2024a) work on integrating generative AI in active learning environments, provides valuable insights into immediate effects but leaves questions about long-term implications largely unexplored. Future longitudinal studies must examine how AI-enhanced learning experiences influence cognitive development, knowledge retention and skill acquisition over extended periods. These studies should specifically investigate how early exposure to AI-assisted learning affects students’ later academic performance and cognitive development.

Kuromiya et al.'s (2020) research on learning analytics provides a foundational framework for measuring long-term educational outcomes. Their work demonstrates the importance of tracking not only academic achievement but also the development of critical thinking skills, creativity and problem-solving abilities. This approach aligns with Williamson's (2024) analysis of AI’s social impact in education, suggesting the need for comprehensive evaluation frameworks that capture both quantitative and qualitative aspects of learning outcomes.

Human–artificial intelligence collaboration models

Research into effective models of collaboration between educators and AI systems represents another crucial area for investigation. Rahiman and Kodikal's (2024) work on AI-empowered learning in higher education reveals complex dynamics in teacher–AI interactions that require further exploration. Their findings suggest that successful integration depends on developing frameworks that leverage the strengths of both human educators and AI systems while maintaining the crucial role of human interaction in education.

Kim's (2024) research on teacher–AI collaboration provides valuable insights into current challenges and opportunities. However, their work also highlights significant gaps in understanding how to optimize these collaborative relationships. Future research must examine how AI can augment rather than replace teacher capabilities, particularly in areas such as personalized instruction, assessment and student support. This aligns with Sposato's (2024) findings on leadership development in the AI era, which emphasize the importance of maintaining human judgement and expertise in educational decision-making.

Cross-cultural and comparative studies

The need for comprehensive cross-cultural analysis of AI’s educational impact emerges from current research gaps identified in the literature. Rappleye et al.'s (2024) work on evidence-based SDG4 discussions demonstrates significant variations in how different cultural contexts approach educational technology integration. Their research reveals the necessity of understanding how cultural factors influence both the acceptance and effectiveness of AI in education.

Wu and Yu's (2024) meta-analysis demonstrates significant variations in how AI implementation outcomes vary across different contexts and cultures. Their findings highlight the need for frameworks that can adapt to diverse educational environments while maintaining consistent ethical standards. This intersection of cultural considerations with governance frameworks points to a critical area requiring systematic investigation: the development of ethical governance structures that can effectively protect student interests across diverse educational contexts.

Ethical governance in artificial intelligence education

The integration of AI in education necessitates a comprehensive ethical framework that addresses governance, privacy and inclusivity as interconnected elements of responsible implementation. Drawing from current research and established principles, this section examines how these components must work together to ensure that AI integration serves educational objectives while protecting student interests and promoting equitable access.

The foundation of ethical AI implementation in education rests on three interconnected pillars: governance structures, privacy protections and inclusive practices. ElSayary's (2024b) work on transformative learning demonstrates how these elements must be considered holistically from the outset of any AI implementation. Their research reveals that effective governance frameworks must balance technological innovation with core educational principles while ensuring robust privacy protections and equitable access. This holistic approach aligns with Williamson's (2024) analysis of AI’s social impact in education, which emphasizes how technological integration affects various aspects of the educational experience simultaneously.

Governance and accountability mechanisms emerge as crucial components in ensuring ethical AI implementation. Lockman and Schirmer's (2020) analysis provides valuable insights into promoting algorithmic fairness and transparency in digital learning environments. Their work demonstrates that effective governance requires clear accountability structures that address both technical performance and educational outcomes. This connects with Brown et al.'s (2024) research on preventing idea “echo-chambers,” which highlights how governance frameworks must actively promote diverse perspectives and critical thinking in AI-enhanced learning environments.

Privacy protection and student autonomy form another critical dimension of ethical AI implementation. Building on Mitchell and Sutherland's (2020) findings on inclusive education, privacy considerations must extend beyond basic data protection to encompass broader questions of student agency and digital rights. Their research reveals how privacy protections particularly affect vulnerable student populations, emphasizing the need for comprehensive protection protocols that address both immediate privacy concerns and longer-term implications for student autonomy. This understanding aligns with Krupar and Taneja's (2020) work on educational rights, which demonstrates how privacy protections serve as fundamental safeguards for student interests while enabling meaningful educational innovation.

The inclusive implementation of AI technologies requires careful attention to equity and accessibility. Rappleye et al.'s (2024) analysis of evidence-based SDG4 discussions demonstrates how implementation frameworks must address both technical and social aspects of accessibility. Their work emphasizes the importance of developing metrics that can evaluate both the effectiveness and equity of AI-enhanced educational practices. These findings suggest that successful ethical frameworks must actively promote inclusive practices while maintaining high standards for educational quality.

As AI technologies continue to evolve, ethical governance frameworks must remain adaptable while preserving core principles of student protection and educational integrity. Terry et al.'s (2024) model for evidence-based practice provides valuable guidance for developing flexible yet robust governance structures. Their work demonstrates how ethical frameworks can evolve with technological advancement while maintaining fundamental educational values and privacy rights. This adaptability proves crucial as educational institutions navigate the rapidly changing landscape of AI technology.

The integration of these ethical considerations requires ongoing evaluation and adjustment. Wu and Yu's (2024) meta-analysis reveals how ethical implementation practices must respond to emerging evidence about AI’s educational impact across different contexts and populations. Their research emphasizes the importance of maintaining ethical standards while adapting to new understanding of AI’s effects on learning processes and educational relationships. This dynamic approach ensures that ethical frameworks remain relevant and effective as AI technologies and educational practices continue to evolve.

Psychological and social impact analysis

Understanding the comprehensive psychological and social effects of AI integration in education represents another critical research priority. Mitchell and Sutherland's (2020) work on special and inclusive education emphasizes the importance of examining how technological interventions affect student motivation, self-efficacy and social-emotional development. Their research suggests the need for detailed investigation of both intended and unintended consequences of AI implementation on educational relationships and social dynamics.

The work of Kelemen and Rumens (2008) on critical management research provides a valuable framework for examining power dynamics and social relationships in AI-enhanced educational environments. Future studies must investigate how AI influences student–teacher relationships, peer interactions and the overall social fabric of educational institutions. This aligns with Riach et al. (2016) analysis of organizational performativity, suggesting the need to examine how AI shapes educational identities and relationships.

Policy development and implementation

Research into effective policy frameworks for AI in education emerges as a crucial priority. Williamson's (2024) analysis of AI’s social life in education demonstrates the complex interplay between technological innovation and educational policy. Their work suggests the need for research that examines how regulatory frameworks can promote responsible AI innovation while protecting educational quality and equity.

Current research by Brown et al. (2024) on preventing idea “echo-chambers” highlights the importance of policy development that promotes diverse perspectives and critical thinking. Future studies must investigate how educational policies can balance the potential benefits of AI with concerns about data privacy, algorithmic bias and educational equity. This aligns with Krupar and Taneja's (2020) work on educational rights, emphasizing the need for policy frameworks that protect student interests while fostering innovation.

Methodological considerations

The complexity of researching AI in education necessitates sophisticated methodological approaches that can capture both quantitative and qualitative aspects of implementation and impact. Drawing from Barends and Rousseau's (2018) work on evidence-based management, several key methodological considerations emerge for future research in this field.

Interdisciplinary collaboration

The examination of AI’s educational impact requires research approaches that integrate perspectives from multiple disciplines. Current work by Pietsch et al. (2024) on open innovation in schools demonstrates the value of collaborative research approaches. Their findings suggest that understanding AI’s educational impact requires expertise from fields including computer science, cognitive psychology, pedagogy and educational theory.

Practical examples of interdisciplinary collaboration

The implementation of effective interdisciplinary research in AI education requires carefully structured collaborative frameworks. Building on Pietsch et al.'s (2024) work on open innovation in schools, several concrete examples illustrate how different disciplines can work together effectively to advance educational AI implementation while maintaining rigorous standards for both innovation and pedagogical effectiveness.

The development and evaluation of adaptive learning systems provides a compelling example of necessary interdisciplinary collaboration in educational AI research. In such projects, computer scientists typically focus on developing core AI algorithms for natural language processing and adaptive response mechanisms, working in close consultation with cognitive psychologists who design assessment metrics for student engagement and cognitive load measurement. Educational researchers simultaneously evaluate the system’s alignment with established learning theories and pedagogical frameworks, while data scientists analyze emerging learning patterns to optimize system performance. Ethics researchers provide crucial oversight regarding privacy considerations and algorithmic bias detection. This collaborative approach ensures that technical innovation serves clear educational objectives while maintaining rigorous standards for student well-being and learning outcomes.

AI-enhanced assessment development represents another area where cross-disciplinary expertise proves essential for effective implementation. Assessment specialists work closely with machine learning experts to develop automated scoring algorithms that align with established evaluation frameworks and rubrics. Educational psychologists play a crucial role in validating assessment outcomes against traditional methods, while subject matter experts ensure content accuracy and standards alignment. Accessibility researchers contribute expertise in inclusive design principles, ensuring that AI-enhanced assessments serve diverse student populations effectively. This integrated approach demonstrates how technical innovation can enhance assessment practices while maintaining pedagogical validity and educational equity.

The implementation of AI support systems in special education contexts further illustrates the importance of diverse expertise in educational technology development. Special education specialists collaborate with user interface designers to develop accessible interfaces that address specific learning needs and accommodation requirements. Speech and language pathologists contribute crucial insights regarding communication support features, while educational technologists ensure seamless integration with existing assistive technologies. Behavioral psychologists analyze student responses and adaptation strategies, providing valuable feedback for system refinement. This collaborative framework ensures that technological innovations effectively serve specific educational needs while maintaining appropriate support structures for vulnerable student populations.

These examples demonstrate how interdisciplinary collaboration can effectively address complex educational challenges while maintaining focus on student needs and learning outcomes. The integration of diverse expertise ensures that AI implementation in education benefits from multiple perspectives while maintaining rigorous standards for both technical innovation and educational effectiveness. As Wu and Yu's (2024) meta-analysis suggests, such collaborative approaches prove particularly valuable in ensuring that AI educational tools serve diverse student populations effectively while promoting equitable access to educational resources.

Research by Terry et al. (2024) on evidence-based practice provides insights into how interdisciplinary approaches can enhance understanding of complex educational phenomena. Future research must develop methodological frameworks that effectively combine insights from different disciplines while maintaining rigorous standards of evidence. This aligns with Owusu-Agyeman's (2024) work on professional development, suggesting the need for methodological approaches that can capture the multifaceted nature of educational innovation.

Innovative research designs

The unique challenges of studying AI in education require the development of sophisticated research designs. Kuromiya et al.'s (2020) work on learning analytics demonstrates the potential of advanced methodological approaches for understanding educational technology impact. Their research suggests the need for designs that can effectively isolate AI’s specific effects while accounting for confounding variables.

ElSayary's (2024a) research on active learning environments provides valuable insights into methodological challenges in studying AI implementation. Future research must develop innovative approaches that can capture both immediate and long-term effects of AI integration while maintaining methodological rigor. This includes the development of AI-specific randomized controlled trials and sophisticated mixed-methods approaches that can provide comprehensive understanding of implementation impacts.

Participatory research approaches

The importance of stakeholder involvement in AI education research emerges from current literature on educational innovation. Riach et al. (2016) work on organizational performativity demonstrates the value of including diverse voices in research design and implementation. Their research suggests that understanding AI’s educational impact requires direct engagement with educators, students and other stakeholders who experience these technologies firsthand.

Data collection and analysis innovation

The complex nature of AI in education requires sophisticated approaches to data collection and analysis. Williamson's (2024) work on the social life of AI in education demonstrates the need for methods that can capture both quantitative metrics and qualitative experiences. Their research suggests the importance of developing new analytical tools that can process and interpret the rich data generated by AI-enhanced learning environments.

The role of educators in artificial intelligence research and implementation

The position of educators in AI integration emerges as a crucial consideration requiring careful examination. Drawing from Sposato's (2024) research on leadership development in the AI era, educators must be understood not merely as implementers of new technology, but as active participants in shaping how AI transforms educational practice. Their work demonstrates that successful AI integration depends significantly on educators’ ability to critically engage with and adapt these technologies to specific pedagogical contexts.

Professional development and adaptation

Kim's (2024) research on teacher–AI collaboration reveals complex dynamics in how educators adapt to and shape AI implementation. Their findings indicate that effective professional development must go beyond technical training to include critical evaluation skills and pedagogical integration strategies. This aligns with Owusu-Agyeman's (2024) analysis of professional development in higher education, which emphasizes the importance of continuous learning and adaptation in response to technological change.

Research by ElSayary (2024a) on active learning environments provides valuable insights into how educators can effectively integrate AI tools into existing pedagogical approaches. Their work demonstrates that successful implementation requires educators to develop new competencies while maintaining their essential role in fostering critical thinking and creativity.

Pedagogical leadership and innovation

The transformation of educational practices through AI implementation requires educators to assume new leadership roles. Pietsch et al.'s (2024) work on open innovation in schools demonstrates how educators can drive meaningful technological integration through collaborative leadership approaches. Their research suggests that educators must be empowered to shape how AI technologies are implemented in their specific educational contexts.

Balancing technology and human connection

The maintenance of meaningful human interaction in AI-enhanced learning environments emerges as a crucial consideration. Mitchell and Sutherland's (2020) work on special and inclusive education demonstrates the continued importance of human connection in effective teaching and learning. Their research suggests that educators must develop strategies for leveraging AI while preserving essential human elements of education.

Conclusion

The integration of AI in education presents both significant opportunities and substantial challenges that require careful consideration and evidence-based approaches. Through systematic analysis of current research and identified gaps in understanding, several key conclusions emerge regarding the future of AI in education.

First, the need for balanced implementation approaches that recognize both opportunities and limitations becomes apparent. Drawing from Rappleye et al.'s (2024) work on evidence-based SDG4 discussions, successful AI integration requires careful attention to both technological capabilities and educational values. Their research demonstrates that effective implementation must balance innovation with preservation of essential educational principles.

Second, the importance of maintaining human-centered educational practices emerges as a crucial consideration. Lockman and Schirmer's (2020) analysis of online instruction provides valuable insights into how technology can enhance rather than replace human interaction in education. Their work emphasizes the need for approaches that leverage AI’s capabilities while preserving the essential human elements of teaching and learning.

Third, the development of comprehensive research agendas emerges as essential for guiding future implementation. Wu and Yu's (2024) meta-analysis of AI chatbots demonstrates the importance of rigorous evaluation in understanding AI’s educational impact. Their work suggests the need for continued investigation of both immediate and long-term effects of AI integration across different educational contexts.

Finally, the alignment of AI implementation with broader educational goals, particularly SDG4, requires ongoing attention and evaluation. Krupar and Taneja's (2020) work on educational rights emphasizes the importance of ensuring that technological innovation serves rather than undermines educational equity and quality. Their research suggests the need for continued focus on how AI can contribute to inclusive and equitable education for all.

Looking forward, the successful integration of AI in education will require sustained commitment to evidence-based practice, ethical implementation, and maintenance of human-centered approaches. As the field continues to evolve, the insights provided by current research must guide the development of implementation strategies that serve educational objectives while addressing potential challenges and limitations.

Final reflections and future directions

Before presenting the complete references, it is important to acknowledge both the limitations and opportunities that emerge from this analysis. Drawing from Kelemen and Rumens' (2008) work on critical research approaches, several key considerations for future development in this field warrant attention.

The implementation of AI in education represents not merely a technological challenge but a fundamental transformation of educational practice. As demonstrated through the analysis of current research, from ElSayary's (2024a) work on active learning environments to Williamson's (2024) examination of AI’s social impact, the successful integration of AI requires careful attention to both technical and human factors.

The journey toward effective AI integration in education demands continued commitment to evidence-based practice while maintaining focus on core educational values. This balance, as emphasized by Mitchell and Sutherland (2020), requires ongoing dialogue between researchers, practitioners and policymakers to ensure that technological innovation serves rather than supersedes educational objectives.

Looking toward the future, several critical areas emerge that will require sustained attention from the educational research community. First, the rapid evolution of AI capabilities necessitates continuous reassessment of implementation strategies. As Wu and Yu's (2024) meta-analysis suggests, the effectiveness of AI tools varies significantly across different educational contexts, indicating the need for more nuanced understanding of how these technologies can best serve diverse learning environments. This understanding must extend beyond mere technical compatibility to encompass cultural, social and pedagogical considerations.

Furthermore, the development of AI in education must be guided by a clear vision of educational purpose. Rappleye et al.'s (2024) work on evidence-based SDG4 discussions highlights how technological innovation must align with broader educational goals and social justice objectives. This alignment becomes increasingly crucial as AI systems become more sophisticated and their potential impact on educational equity grows more significant. Future research must carefully examine how AI implementation affects different student populations and work to ensure that technological advancement serves to reduce rather than exacerbate educational disparities.

The role of educators in shaping AI integration emerges as another critical consideration for future development. Brown et al.'s (2024) research on preventing idea “echo-chambers” demonstrates the essential role of human judgment in guiding technological implementation. As AI systems continue to evolve, maintaining the proper balance between automated and human-led instruction will require ongoing attention and adjustment. This balance must be informed by careful research into how different combinations of human and AI instruction affect student learning outcomes, engagement and development of critical thinking skills.

Additionally, the ethical dimensions of AI in education will require continued scrutiny and development. Building on Krupar and Taneja's (2020) work on educational rights, future research must address emerging ethical challenges related to privacy, autonomy and equity in AI-enhanced learning environments. This includes developing more sophisticated frameworks for understanding how AI systems influence student development and learning processes, as well as establishing clear guidelines for responsible implementation that protects student interests while promoting educational innovation.

The path forward requires a delicate balance between embracing AI’s potential and maintaining critical awareness of its limitations. As Kim's (2024) research on teacher–AI collaboration suggests, successful integration depends on developing approaches that leverage technological capabilities while preserving the essential human elements of education. This understanding must guide future research and implementation efforts, ensuring that AI serves as a tool for enhancing rather than replacing effective educational practices.