Psychological Impacts of AI Use in Schools

In addition to the focus on teaching and learning, there are other attributes that can affect student success. With this in mind, I would like to share this recent article entitled, “Psychological impacts of AI use on school students: A systematic scoping review of the empirical literature” by Kundu and Bej (2025).

The authors conducted a systematic scoping review to explore how these technologies influence students’ psychological, cognitive, emotional, and behavioral development across preschool to high school. Their analysis of 24 empirical studies (2019–2023) across nine countries offers one of the most comprehensive portraits to date of how AI supports and sometimes strains how people learn.

Findings

• Increased engagement, cognitive achievement, self-efficacy, learning autonomy, and decreased frustration; and 

• Over reliance, anxiety, stress, social isolation, unstable mental health, privacy, bias, and justice.

Across contexts, AI demonstrated significant potential to enhance cognitive functioning, especially when used for personalized learning, immediate feedback, and simulation-based environments. Studies reviewed (Weng et al., 2024) show that AI improved critical thinking, problem-solving, and computational efficacy. Cognitive benefits often arose through mechanisms aligned with constructivist and information processing theories students learned best when AI tools scaffolded challenges and provided feedback at their individual zone of proximal development (Vygotsky, 1978; Bandura, 1977).

However, cognitive overload and overreliance on automation emerged as recurring limitations.

The review’s emotional dimension revealed that AI can foster both empowerment and anxiety. Conversational AI agents' pedagogical tools encouraged curiosity and self-expression, while AI-supported therapy chatbots improved adolescent well-being (Vertsberger et al., 2022).

Yet, affective costs were also observed. AI’s impersonality sometimes led to diminished emotional perception and reduced empathy in peer or teacher interactions (Lai et al., 2024). Students reported discomfort with algorithmic judgment and emotional mimicry, suggesting a need for more emotionally intelligent AI that sustains authentic human connection.

Behaviorally, AI exposure reshaped students’ study habits and communication patterns. Learners displayed greater engagement, persistence, and collaboration, particularly when AI systems were gamified or interactive (Sanusi et al., 2023; Lin et al., 2020). However, the review cautioned against technology dependence and social withdrawal, particularly among adolescents immersed in AI-based environments that replace rather than augment social interaction.

Kundu and Bej’s SWOT synthesis underscored that AI’s strengths in engagement and personalization must be balanced against threats of addiction, inequity, and data privacy. Across age levels, benefits were most consistent when AI supplemented rather than supplanted teacher-led pedagogy, reinforcing that effective AI integration should always be human-centered.

The study’s SoTL implications align with decades of research on how humans learn effectively. Learning gains occur when AI systems support active cognition, social interaction, metacognition, and emotional regulation core principles from Piagetian constructivism, Bruner’s discovery learning, and contemporary cognitive science. For educators, the key takeaway is not to automate learning but to design AI-mediated environments that amplify human curiosity, empathy, and reflective thinking.

Kundu and Bej conclude that while AI can enhance academic achievement and emotional resilience, it also introduces psychological vulnerabilities from identity anxiety to cognitive overload. Their findings support a balanced, ethical, and developmentally appropriate approach to AI in education. Teachers and policymakers must integrate AI tools within safe, relational, and inclusive learning ecologies, ensuring students develop both digital competence and socio-emotional literacy.

References

Bandura, A. (1977). Social learning theory. Prentice Hall.

Chiu, T. K. F., Lo, M. F., & Lai, C. (2022). Examining overreliance on automation in AI-assisted learning environments. British Journal of Educational Technology, 53(6), 1590–1605. https://doi.org/10.1111/bjet.13238 

Kundu, A., & Bej, T. (2025). Psychological impacts of AI use on school students: A systematic scoping review of the empirical literature. Research and Practice in Technology Enhanced Learning, 20(30). https://doi.org/10.1186/s41078-025-00330-9 

Lai, J., Lin, X., & Zhang, Y. (2024). Emotional dissonance in AI-assisted classrooms: Implications for empathy and social learning. Computers in Human Behavior, 156, 108472. https://doi.org/10.1016/j.chb.2024.108472 

Lin, T., Hung, C., & Chen, W. (2020). Gamified AI-based learning environments and student collaboration in STEM education. Interactive Learning Environments, 28(8), 1018–1034. https://doi.org/10.1080/10494820.2019.1579232 

Sanusi, I., Tan, J., & Wong, M. (2023). Game-based AI learning for improved collaboration and engagement: A mixed-methods analysis. Computers & Education Open, 5, 100145. https://doi.org/10.1016/j.caeo.2023.100145 

Vertsberger, D., Cohen, A., & Barak, M. (2022). Supporting adolescents’ emotional well-being using AI-enabled chatbots in educational settings. Computers in Human Behavior, 137, 107410. https://doi.org/10.1016/j.chb.2022.107410 

Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Harvard University Press.

Weng, J., Xie, Y., & Zhai, X. (2024). Integrating generative AI into project-based learning: Effects on cognitive engagement and problem-solving. Computers and Education: Artificial Intelligence, 6, 100231. https://doi.org/10.1016/j.caeai.2024.100231