In an era increasingly dominated by digital technologies and artificial intelligence, new research from Chen, Chen, Yang, and colleagues delves deeply into the complex psychological dynamics influencing university students’ relationship with generative artificial intelligence (GAI). Published in the International Journal of Mental Health and Addiction, the study offers a nuanced exploration of how self-esteem, academic anxiety, and the perceived usefulness of AI tools collectively shape GAI dependency among young adults navigating higher education.
The study confronts a timely and emergent concern: the extent to which university students rely on generative AI platforms for academic tasks, and how this dependency is intertwined with their internal psychological landscape. With AI tools offering unprecedented access to information and assistance in academic work, students’ interaction with these systems is not merely functional but also deeply psychological, implicating factors such as self-worth and emotional wellbeing.
At the heart of the research lies self-esteem, a crucial personal attribute often defined as one’s overall evaluation of self-worth. The investigators propose that self-esteem may influence, directly or indirectly, the degree to which students engage with GAI technology. Individuals with lower self-esteem might be more prone to increased dependency, seeking validation or assistance that compensates for perceived academic inadequacies or anxieties.
The authors foreground academic anxiety as a pivotal mediating variable in this relationship. Academic anxiety encompasses the apprehensions and stressors specific to academic performance and environments, which can severely impact cognitive functioning and motivation. This anxiety might push students toward overreliance on AI tools as coping mechanisms, engendering a paradox where reliance intended to alleviate stress may trap them in a cycle of dependency.
Another essential mediator examined is perceived usefulness—students’ subjective evaluation of how beneficial they find AI tools for achieving academic goals. The perception of AI as an indispensable study aid could potentiate reliance, especially if students believe these tools can enhance productivity, understanding, or grades. Hence, perceived usefulness offers a motivational factor that shapes dependency behavior.
Employing a multiple mediation model, the researchers undertake a sophisticated analytical approach to disentangle the layered pathways linking self-esteem to GAI dependency. Their methodology allows for a detailed characterization of both direct and indirect effects, illuminating how academic anxiety and perceived usefulness intervene in this psychological interplay. Through this approach, the study advances a comprehensive understanding beyond simplistic cause-effect assumptions.
The results reveal that self-esteem negatively correlates with GAI dependency, indicating that students with higher self-esteem tend to exhibit a more balanced relationship with AI assistance. However, this association is significantly mediated by academic anxiety and perceived usefulness, such that lower self-esteem leads to heightened academic anxiety, which in turn increases perceived usefulness of AI tools, culminating in stronger dependency behaviors.
Intriguingly, the study suggests a dual mediating pathway: academic anxiety exacerbates the perceived indispensability of AI, suggesting that anxiety acts as an emotional driver behind the practical evaluation of AI usefulness. This nuanced finding demonstrates that psychological distress does not operate independently but colors the cognitive appraisal of technological aids, amplifying reliance.
From a theoretical standpoint, this research integrates concepts from self-determination theory and cognitive appraisal frameworks to thoughtfully interpret how students internalize and react to technological environments. The authors argue that interventions aimed at reducing academic anxiety or enhancing self-esteem could attenuate problematic AI reliance, fostering healthier engagement with digital tools.
In practical terms, the study underscores the importance for educators and policymakers to acknowledge the psychological dimensions shaping technology adoption in academic contexts. Rather than dismissing AI engagement as mere convenience or cheating, it advocates for a supportive approach recognizing underlying mental health factors and guiding constructive usage.
Furthermore, by mapping the mediating factors, this research contributes valuable insights for developers of AI educational platforms. Designing environments that are transparent, supportive, and calibrated to reduce academic stress could mitigate the formation of unhealthy dependency patterns and promote autonomous learning.
The research also raises critical ethical concerns regarding the fine line between beneficial AI assistance and overdependence that might undermine skill development or exacerbate anxiety. Understanding students’ psychological profiles in tandem with their technology use highlights the need for balanced integration that empowers rather than disables learners.
Looking ahead, the authors recommend longitudinal studies to track changes in self-esteem, anxiety, and AI use over time, enabling a dynamic understanding of how these relationships evolve during students’ academic careers. Such insights would be vital for tailoring interventions responsive to developmental trajectories.
Although the study provides compelling evidence from a substantial sample of university students, it also calls for cross-cultural investigations to ascertain the generalizability of these findings. Sociocultural factors may differentially influence self-esteem constructs, academic anxiety prevalence, and technology adoption patterns, presenting a fertile domain for future research.
In sum, the work by Chen et al. marks a significant advancement in mental health and technology addiction literature by unveiling the complex psychological mediators critical to understanding AI dependency in academic settings. Their integrative approach offers both theoretical contributions and practical applications essential for navigating the interface of human psychology and artificial intelligence.
As generative AI continues to embed itself into educational paradigms, comprehending the intricate mechanisms that govern student interaction with such technology is imperative. This research paves the way toward interventions and policies that responsibly harness AI’s potential while safeguarding mental health and promoting holistic academic growth.
Subject of Research: University students’ psychological factors affecting dependence on generative artificial intelligence in academic settings.
Article Title: University Students’ Self-Esteem and GAI Dependency: Multiple Mediating Roles of Academic Anxiety and Perceived Usefulness.
Article References:
Chen, Y., Chen, T., Yang, S. et al. University Students’ Self-Esteem and GAI Dependency: Multiple Mediating Roles of Academic Anxiety and Perceived Usefulness. Int J Ment Health Addiction (2025). https://doi.org/10.1007/s11469-025-01581-4
Image Credits: AI Generated
DOI: https://doi.org/10.1007/s11469-025-01581-4
