آگاهی تهدید محور از هوش مصنوعی و کناره گیری شغلی معلمان: نقش میانجی نشخوار فکری و خستگی عاطفی

Purpose: The rapid advancement of emerging technologies, particularly artificial intelligence (AI), has profoundly transformed traditional work structures and professional environments. In the education sector, while AI offers significant opportunities to enhance teaching and learning processes, it has simultaneously generated substantial concerns among teachers regarding job security, professional identity, and future career prospects. Existing literature has largely focused on the technological benefits of AI, with limited attention given to its psychological and behavioral consequences. This gap is particularly critical in private educational settings, where job insecurity and competitive pressures are more pronounced. Therefore, a significant research gap exists in understanding the psychological mechanisms linking AI awareness to teachers’ work-related behaviors. This study aims to investigate the impact of threat-based AI awareness on teachers’ work withdrawal and to examine the mediating roles of negative work-related rumination and emotional exhaustion within the framework of Conservation of Resources (COR) theory. The study seeks to answer whether AI awareness can indirectly increase teachers’ withdrawal behaviors through cognitive and emotional processes.
Methodology: This study is applied in nature and uses a quantitative, descriptive-correlational design with an inductive approach. The theoretical framework is based on COR theory, which emphasizes the role of perceived threats to psychological resources in shaping behavioral responses. The statistical population consisted of teachers from non-governmental elementary schools in Mashhad, Iran. A sample of 200 participants was selected using stratified random sampling. Data were collected using a standardized questionnaire consisting of 29 items measuring four constructs: awareness of AI, negative work-related rumination, emotional exhaustion, and disengagement from work. Validity was confirmed by confirmatory factor analysis and reliability was confirmed using Cronbach's alpha and composite reliability. Data analysis was performed using partial least squares structural equation modeling (PLS-SEM) with SmartPLS4 software. The model fit was evaluated with indices such as SRMR, NFI, and R², and hypothesis testing was performed using bootstrap resampling’s.
Results: The results obtained from the structural equation modeling analysis indicate that the proposed research model demonstrates an acceptable and robust level of fit with the empirical data. Key model fit indices, including the Standardized Root Mean Square Residual (SRMR = 0.061) and the Normed Fit Index (NFI = 0.91), confirm the adequacy of the model’s overall structure. In addition, the coefficients of determination (R²) for the endogenous constructs reveal substantial explanatory power. Specifically, the model explains 89% of the variance in work withdrawal (R² = 0.89), 82.6% of the variance in emotional exhaustion, and 77.6% of the variance in negative work-related rumination. These high explanatory values indicate that the model provides a strong and reliable representation of the relationships among the studied variables.
At the level of direct effects, the findings show that threat-based AI awareness has a positive and statistically significant impact on teachers’ work withdrawal (β = 0.299, t = 4.729). This result suggests that as teachers increasingly perceive artificial intelligence as a threat to their professional roles and job security, their tendency to engage in withdrawal behaviors—such as reduced participation, psychological disengagement, and diminished effort—significantly increases. Furthermore, AI awareness exerts a very strong and significant effect on emotional exhaustion (β = 0.909, t = 92.598), highlighting its critical role as a major psychological stressor that depletes teachers’ emotional and cognitive resources. Emotional exhaustion, in turn, is found to have a positive and significant effect on work withdrawal (β = 0.287, t = 3.878), confirming its role as a key predictor of disengagement-related behaviors.
In terms of cognitive pathways, the results demonstrate that AI awareness has a strong and significant positive effect on negative work-related rumination (β = 0.881, t = 67.469). This finding indicates that heightened awareness of AI-related threats leads to persistent and repetitive negative thinking about work-related issues, particularly concerning job insecurity and professional identity. Additionally, negative rumination significantly predicts work withdrawal (β = 0.390, t = 6.014), suggesting that continuous cognitive preoccupation with work-related stressors prevents psychological recovery and contributes to behavioral disengagement.
At the level of indirect effects, the findings confirm that both emotional exhaustion and negative rumination serve as significant mediating mechanisms in the relationship between AI awareness and work withdrawal. The indirect effect of AI awareness on work withdrawal through emotional exhaustion is estimated at 0.260 (t = 3.625), while the indirect effect through negative rumination is stronger, at 0.343 (t = 5.122). This comparison indicates that cognitive processes (i.e., rumination) play a more dominant role than emotional processes in transmitting the effects of perceived AI threats to withdrawal behaviors. In other words, the way individuals cognitively process and internalize perceived threats appears to be more influential than emotional depletion alone in shaping behavioral outcomes.
Overall, the pattern of results suggests that AI awareness influences work withdrawal not only directly but also indirectly through simultaneous cognitive and emotional mechanisms. These findings point to a cascading resource depletion process, in which perceived technological threats activate negative cognitive cycles and emotional strain, ultimately leading to reduced work engagement and increased withdrawal behaviors. This integrated mechanism highlights the complex interplay between cognition and emotion in shaping employees’ responses to technological change. It also underscores the importance of considering both psychological dimensions when examining the organizational consequences of emerging technologies such as artificial intelligence.
Discussion: Within the framework of COR theory, the findings suggest that AI awareness functions as a job stressor that threatens teachers’ psychological resources, triggering cognitive (rumination) and emotional (exhaustion) responses. These processes lead to resource depletion, ultimately resulting in withdrawal behaviors as a protective mechanism. The results are consistent with prior research indicating that perceived technological threats can increase anxiety, reduce motivation, and lead to disengagement. Theoretically, this study contributes by integrating cognitive and emotional mediators into a comprehensive model explaining AI-related behavioral outcomes. Practically, the findings emphasize the importance of addressing the psychological dimensions of digital transformation in educational settings.
Conclusion: This study demonstrates that threat-based AI awareness, in the absence of adequate organizational and psychological support, can undermine teachers’ job stability and knowledge sustainability. By identifying the mediating roles of rumination and emotional exhaustion, the research provides a deeper understanding of how AI-related perceptions influence work behaviors. A key contribution is the development of an integrated model for analyzing the psychological impacts of AI in educational contexts. However, limitations such as the focus on a single city and occupational group restrict generalizability. Future research is recommended to adopt longitudinal designs and explore diverse educational and organizational settings. Ultimately, the study underscores the necessity of implementing supportive interventions, enhancing technological literacy, and strengthening teachers’ psychological resilience in the face of emerging technologies.