Abstract: As the core driving force of the Fourth Industrial Revolution, Artificial Intelligence (AI)has reshaped organizational models and transformed the way employees work. As AI application in the workplace continues to deepen, it profoundly impacts employees' psychology and behavior. However, there is still a relatively scarce body of academic research on the consequences of AI usage, and the conclusions are mixed. Some scholars argue that AI usage can lead to positive effects, such as increased job engagement and job autonomy, while others contend that AI usage can result in negative effects, such as increased job insecurity and job burnout. While existing studies have explored the potential positive or negative impacts of AI usage, there is a lack of integrated analysis of these dual effects. Additionally, most studies have primarily examined the mechanisms through which AI usage affects job engagement and proactive service behaviors, with relatively few focusing on its impact on employee innovation performance. Employee innovation performance is crucial for organizations to gain a competitive advantage and achieve high-quality development in the complex and changing business environment of the digital intelligence age. Therefore, it is necessary to further explore the relationship between AI usage and employee innovation performance. Based on the stress cognitive appraisal theory, this study empirically analyzes the double-edged sword effect of AI usage on employee innovation performance, using cognitive appraisal as a mediator and leaders' AI symbolization as a moderator.
Data are collected through a questionnaire with samples from the “Credamo” online platform. To minimize the effect of common method variance, this study adopts a three-wave research design, with data collected at three different time points. Ultimately, a total of 387 valid samples are obtained by matching longitudinal data. This study uses Mplus 8. 0 for confirmatory factor analysis of the variables involved and SPSS 25. 0 for descriptive statistics, correlation analysis, and moderating effects analysis. Additionally, mediation effects and moderated mediation effects tests are conducted using the Process macro program in SPSS 25.0 with the bootstrap method.
This study finds that AI usage has a double-edged sword effect on employee innovation performance. Specifically, AI usage reduces employee innovation performance through threat appraisal and increases employee innovation performance through challenge appraisals. Threat and challenge appraisal partially mediate the relationship between AI usage and employee innovation performance. Leaders' AI symbolization weakens the positive correlation between AI usage and threat appraisal, as well as the mediating effect of threat appraisal in the relationship between AI usage and employee innovation performance. Meanwhile, leaders' AI symbolization strengthens the positive relationship between AI usage and challenge appraisal, as well as the mediating effect of challenge appraisal in the relationship between AI usage and employee innovation performance.
This study has several theoretical implications. First, it enriches the research on the relationship between AI usage and employee innovation performance, confirms the double-edged sword effect of AI usage on employee innovation performance, and helps resolve previous contradictory arguments about the consequences of AI influence. Second, based on the stress cognitive appraisal theory, this study constructs a two-path model of the impact of AI usage on innovative performance, reveals the mechanism of AI usage influencing employee innovation performance, and enriches the research on the mediating mechanism of AI usage. Finally, the moderating role of leaders' AI symbolization is examined and verified, expanding the boundary conditions under which AI usage affects employee work outcomes and enriching empirical research on leaders' AI symbolization.
This study also provides practical implications for managing employee innovation in digital workplaces. Organizations and employees should rationally understand and correctly approach the use of AI, recognizing it as a double-edged sword for employee innovation performance. Efforts should be made to alleviate the negative effects of AI usage while leveraging its positive impact. Organizations should guide employees in developing a correct perception of AI and ensure that AI plays a positive role in the workplace. Additionally, organizations should focus on leadership development and enhance the AI symbolization behavior of managers to promote employee innovation performance.

Key words: AI Usage, Threat Appraisal, Challenge Appraisal, Leaders' AI Symbolization, Employee Innovation Performance

摘要： 在数智化时代,人工智能在工作场所的应用不断深化,其对员工创新绩效的影响仍需深入探索。基于压力认知评估理论,本研究通过对387份问卷多时间节点的数据进行分析,结果发现,人工智能使用会通过威胁性评估抑制员工的创新绩效,但会通过挑战性评估提升员工的创新绩效,威胁性评估和挑战性评估在人工智能使用对员工创新绩效的影响中起中介作用;领导AI符号化不仅调节人工智能使用与两类认知评估之间的关系强度,还显著调节人工智能使用通过两类认知评估影响员工创新绩效的间接作用,但当领导AI符号化程度高时,挑战性评估的中介作用更强,而威胁性评估的作用更弱。研究结论揭示了人工智能使用对员工创新的双刃剑效应,为智能化工作场域中的员工创新管理提供了实践启示。

关键词: 人工智能使用, 威胁性评估, 挑战性评估, 领导AI符号化, 创新绩效