The First Scene of Artificial Intelligence Curriculum Reform: Longitudinal Evidence on Teachers’ Stress and Multidimensional Anxiety

Against the backdrop of a global push for digital transformation in education, the formal integration of artificial intelligence (AI) courses into basic education curricula presents a significant challenge for educators. A recent longitudinal study has shed light on the psychological pressures teachers face during this critical reform period, revealing a sustained increase in stress and anxiety, and highlighting the crucial role of school support in mitigating these effects. The research, conducted over two waves in late 2025, involved 553 primary and secondary school teachers, with 348 providing matched data for the follow-up assessment.

The findings underscore a pressing concern for educational policymakers and administrators: the early phase of AI curriculum reform is not a fleeting period of adjustment but a sustained psychological risk for teachers. Without timely and targeted interventions, this pressure could escalate into significant teaching anxiety, potentially impacting educational quality and teacher retention.

AI Integration Creates Rising Stress and Anxiety for Educators
The study’s initial findings, published in Frontiers in Psychology, reveal a clear upward trend in both perceived AI teaching stress and AI teaching anxiety among teachers from October to December 2025. The mean score for perceived AI teaching stress rose from 3.685 to 3.879, while AI teaching anxiety increased from 3.596 to 3.924. This progression was consistent across all measured dimensions of anxiety: skill, replacement, ethical, and adaptation anxiety. Notably, replacement anxiety saw the most significant increase, suggesting that teachers are increasingly concerned about the potential impact of AI on their professional roles and value.

This sustained increase in stress and anxiety is attributed to the multifaceted demands placed upon educators. The AI curriculum reform requires teachers not only to master new technical skills but also to grasp complex AI principles, redesign pedagogical approaches, navigate ethical considerations, and ensure data privacy and safety. This comprehensive shift goes far beyond traditional technology integration training, demanding a significant investment of time, energy, and cognitive resources.

Longitudinal analysis further revealed that stress and anxiety at the first time point (T1) were significant predictors of stress and anxiety at the second time point (T2). Specifically, perceived AI teaching stress at T1 significantly predicted perceived AI teaching stress at T2 (path coefficient of 0.521), indicating a persistent experience of pressure. Similarly, T1 AI teaching anxiety was a strong predictor of T2 AI teaching anxiety (path coefficient of 0.463), demonstrating the cross-time continuity of these psychological burdens. These findings support the hypothesis that the challenges posed by AI curriculum reform are not temporary but tend to endure, requiring ongoing attention and support.

The research team employed a two-wave longitudinal design to capture these dynamic changes. The first wave, conducted in October 2025, captured the initial impact of the reform as it entered school teaching practices. The second wave, in December 2025, assessed teachers’ psychological states after they had experienced a period of implementation, allowing for the examination of temporal stability and the evolution of stress and anxiety.

The study’s theoretical framework draws upon Conservation of Resources Theory, Social Support Theory, and Ecological Systems Theory. Conservation of Resources Theory posits that individuals experience stress when they perceive a threat to their resources, actual resource loss, or when resource investment yields insufficient returns. In the context of AI reform, teachers are investing significant resources (time, effort, cognitive load) without immediate or certain returns, leading to resource depletion. Social Support Theory highlights the potential for school support to buffer the negative effects of stress. Ecological Systems Theory emphasizes that individual development and adaptation are influenced by the microecological environment, suggesting that differences in school levels (primary, junior secondary, senior secondary) may lead to varied experiences of stress and anxiety.

The study’s methodology involved a two-wave longitudinal survey of 553 primary and secondary school teachers in Beijing, Shanghai, Guangzhou, and Hangzhou – cities at the forefront of AI curriculum reform in China. Following data quality control and attrition analysis, 348 teachers provided matched data for both waves. The measurement instruments, adapted from internationally recognized scales, assessed perceived AI teaching stress, AI teaching anxiety (comprising skill, replacement, ethical, and adaptation dimensions), and school support (including emotional, resource, training, and institutional support). Qualitative interviews with ten teachers further enriched the quantitative findings, providing deeper insights into their lived experiences.

School Support as a Critical Buffer
A key finding of the study is the significant buffering role of school support in mitigating the relationship between stress and anxiety. While school support did not eliminate the source of stress, it demonstrably weakened the transmission of stress into anxiety. This aligns with the buffering hypothesis of social support, which suggests that support mechanisms help reduce the negative impact of stressors.

The research identified that emotional support played a consistent role in weakening the longitudinal continuity of perceived AI teaching stress, particularly in relation to skill, ethical, and replacement anxiety. This suggests that a sense of psychological safety, understanding, and recognition from school leaders and peers is fundamental for teachers navigating the uncertainties of AI reform.

Training support also emerged as a significant buffer, particularly in reducing the association between stress and skill and replacement anxiety. This highlights the importance of providing targeted and practical training that directly addresses teachers’ competence gaps and reassures them about their professional roles in an AI-integrated educational landscape. Resource support demonstrated a buffering effect on adaptation and replacement anxiety, indicating that access to appropriate tools, materials, and practical assistance can alleviate concerns about classroom implementation and job security. Institutional support was found to significantly buffer ethical anxiety and replacement anxiety, underscoring the need for clear policies, guidelines, and evaluation frameworks to reduce ambiguity and uncertainty.

The qualitative data corroborated these quantitative findings. Teachers frequently cited a lack of adequate training, insufficient resources, and institutional misalignment as factors exacerbating their anxiety. Conversely, positive feedback, peer support, in-depth training, and collaborative learning experiences were identified as crucial alleviating factors. This qualitative evidence provides a human-centered perspective on the statistical relationships, illustrating how concrete support mechanisms translate into reduced psychological distress.

Implications for Educational Policy and Practice
The study’s findings carry significant implications for educational policy and practice, particularly concerning the timing and targeting of support interventions. The research strongly suggests that the early phase of AI curriculum reform is a critical window for implementing supportive measures. Proactive and staged interventions are crucial, rather than reactive responses after teachers have already experienced burnout.

One key recommendation is to prioritize emotional support in the initial stages of reform. This could involve school leaders publicly acknowledging the challenges, fostering a culture of tolerance for experimentation, and establishing regular channels for teachers to share experiences and concerns. Peer support networks can also play a vital role in creating a sense of community and shared problem-solving.

Furthermore, the study emphasizes the need for targeted training and institutional support. Training should move beyond basic technical skills to encompass pedagogical integration and address teachers’ specific anxieties about job replacement and ethical responsibilities. Clear institutional guidelines on data privacy, ethical use of AI, and assessment frameworks are essential to reduce uncertainty and provide a stable operational environment.

The research also highlights the importance of tailoring support strategies to different school levels. Primary school teachers, for instance, may require more support related to classroom adaptation and resource availability, while secondary school teachers might benefit more from clearer ethical guidelines and institutional frameworks to address concerns about professional value and evaluation.

Limitations and Future Directions
While this study offers valuable insights, it also presents certain limitations. The two-wave longitudinal design, while informative, primarily establishes associations and continuity rather than definitive causal links. Future research could benefit from more robust designs, such as quasi-experimental or more extensive longitudinal tracking, to further elucidate causal mechanisms. The reliance on self-reported data also introduces the potential for social desirability bias, suggesting that future studies could incorporate multi-source data, including classroom observations and administrative records, to provide a more comprehensive picture. Additionally, the sample was drawn from regions at the forefront of AI reform; extending the research to other geographical areas and later stages of reform would enhance the generalizability of the findings.

Conclusion
In conclusion, the integration of AI into basic education is a complex process that significantly impacts teachers’ psychological well-being. This study provides robust evidence that the early phase of AI curriculum reform is characterized by rising stress and anxiety, with significant cross-time continuity. The research underscores the multidimensional nature of teacher anxiety and highlights the critical, albeit conditional, buffering role of school support. By offering timely, targeted, and context-specific support, educational institutions can help teachers navigate the challenges of AI integration, fostering a more positive and effective learning environment for all. The findings serve as a crucial call to action for policymakers and school administrators to prioritize teacher well-being as a cornerstone of successful educational reform.