AI Competence and Employee Well-Being : A Moderated Mediation Model of Job Crafting and Autonomy

다국어 초록 (Multilingual Abstract)

As artificial intelligence (AI) continues to expand across workplace settings, employees are increasingly expected to work alongside AI systems when carrying out their job responsibilities. Rather than functioning solely as a supporting technology, AI now contributes to information processing, service provision, communication activities, and various forms of work-related decision making. Consequently, employees are required not only to adapt to technological changes but also to develop the capability to incorporate AI effectively into their daily work practices. Although organizations are investing heavily in AI-enabled systems, much of what has been studied so far centers on questions of technology adoption, system effectiveness, and organizational performance. Yet what employees actually feel and experience as AI becomes part of their daily work has received far less scrutiny. In particular, limited understanding exists regarding how employees perceive their own capability to work with AI and how such perceptions influence subsequent work behaviors and employee well-being.
This study turns to SDT to ask a more specific question: when employees feel capable of working with AI, does that sense of capability actually translate into greater well-being, and if so, through what process? For the purposes of this study, AI competence refers to how capable employees feel when it comes to putting AI technologies to effective use in carrying out their work tasks. Moving beyond perspectives that emphasize technological characteristics alone, this study focuses on employees’ subjective capability beliefs as an important psychological resource in AI-enabled workplaces. Employees with stronger AI competence are, in this view, more inclined to play an active part in determining how their work takes shape, adjusting tasks, seeking out better fits, and reorganizing how they spend their time. Over time, these small but sustained changes in how work is done are what drive improvements in well-being. Accordingly, job crafting is conceptualized as a key mechanism through which AI competence exerts its influence on employee well-being. In this view, capability beliefs generate value not by their mere presence but by motivating employees to enact meaningful changes in the way they perform their work.
The study further argues that the translation of AI competence into proactive behavior may depend on contextual conditions within the work environment. Drawing upon the autonomy perspective within SDT, autonomy is proposed as a facilitating condition that enables employees to act on their capability beliefs. When employees feel they have genuine input into how their work is structured, they are in a stronger position to draw on their AI-related confidence, making it more likely that they will take steps to reshape their tasks. On the flip side, when employees have little control over their work, high AI competence alone is not enough to get them to actively reshape their tasks. This is why autonomy matters here: it shapes how strongly AI competence connects to job crafting, and by extension, how much that connection ends up benefiting well-being.
To check whether these relationships hold up, the study collected survey data from service-sector employees across three separate rounds of data collection. AI competence was collected from employees at Time 1. Four weeks later, at Time 2, they answered questions about job crafting and autonomy. Another four weeks later, at Time 3, well-being was assessed. Confirmatory factor analysis, hierarchical regression, and bootstrapping were the primary analytical tools used to evaluate the hypotheses. Spreading the measurements across three time points helped keep the variables separate and reduced the risk of common method bias.
Employees who felt more capable with AI were indeed more likely to take an active role in reshaping their work, supporting the predicted link between AI competence and job crafting. Job crafting in turn predicted better well-being. When job crafting entered the model, the direct path from AI competence to well-being dropped out, which points to job crafting as the key mechanism connecting the two. The moderating role of autonomy was also confirmed: where workplace discretion was higher, the relationship between AI competence and job crafting grew stronger. When autonomy was high, capability beliefs were more readily converted into concrete, proactive changes to how work was designed and carried out. By extension, the well-being benefits that AI competence produced through job crafting were more pronounced when autonomy was high.
Overall, the findings suggest that the organizational benefits associated with AI do not arise solely from technological implementation. What seems to matter more is whether employees feel confident using AI at work and whether that confidence actually shows up in how they go about their daily tasks. In doing so, the study extends SDT into AI-enabled work settings, showing not just that AI competence matters for employee well-being, but how and when it does. Capability beliefs, proactive behavior, and work context all play connected roles in this process, and taken together they paint a clearer picture of how employees adapt as AI becomes a bigger part of working life. For organizations navigating AI-driven change, the findings offer some concrete directions for supporting employee well-being along the way.

목차 (Table of Contents)

• CHAPTER 1. Introduction 1
• 1.1 Research Question 1
• 1.2 Research Gap 4
• 1.3 Research Objectives 7
• 1.4 Structure of the Dissertation 9
• CHAPTER 2. Literature Review 12
• 2.1 Theoretical Foundations 12
• 2.1.1 Self-Determination Theory (SDT) 12
• 2.1.2 Application of SDT in an Organizational Context 14
• 2.2 AI Competence 17
• 2.2.1 Definition of AI Competence 17
• 2.2.2 Literature Review on AI Competence 21
• 2.3 Employee Well-Being 24
• 2.3.1 Definition of Employee Well-Being 24
• 2.3.2 Literature Review on Employee Well-Being 27
• 2.4 Job Crafting 30
• 2.4.1 Definition of Job Crafting 30
• 2.4.2 Literature Review on Job Crafting 34
• 2.5 Autonomy 37
• 2.5.1 Definition of Autonomy 37
• 2.5.2 Literature Review on Autonomy 40
• CHAPTER 3. Research Model and Hypotheses 45
• 3.1 Hypothesis Model 45
• 3.2 Hypothesis Development 46
• 3.2.1 AI Competence and Employee Well-Being 46
• 3.2.2 AI Competence and Job Crafting 50
• 3.2.3 The Mediating Role of Job Crafting 53
• 3.2.4 The Moderating Role of Autonomy 57
• 3.2.5 The Moderated Mediation Effect 60
• CHAPTER 4. Methods 63
• 4.1 Research Design and Procedure 63
• 4.2 Sample 64
• 4.3 Measures 70
• 4.3.1 AI Competence 70
• 4.3.2 Autonomy 71
• 4.3.3 Job Crafting 71
• 4.3.4 Employee Well-Being 72
• 4.3.5 Control Variables 72
• 4.4 Analytical Methods 73
• 4.4.1 Validity and Reliability Testing 73
• 4.4.2 Hypothesis Testing Procedures 73
• 4.4.3 Common Method Bias 74
• CHAPTER 5. Results 80
• 5.1 Preliminary Analyses 80
• 5.1.1 Confirmatory Factor Analysis and Validity Check 80
• 5.1.2 Descriptive Statistics and Correlations 84
• 5.2 Hypothesis Testing 86
• 5.2.1 Test of Hypothesis 1 87
• 5.2.2 Test of Hypothesis 2 88
• 5.2.3 Test of Hypothesis 3 90
• 5.2.4 Test of Hypothesis 4 91
• 5.2.5 Test of Hypothesis 5 93
• 5.2.6 Summary of Hypotheses Testing Results 94
• CHAPTER 6. Discussion 96
• 6.1 Summary of Findings 96
• 6.2 Implications 100
• 6.2.1 Theoretical Implications 100
• 6.2.2 Practical Implications 104
• 6.3 Limitations and Future Research 109
• References 115
• Appendix 136
• 1. Supplementary Analysis 136
• 2. Survey in English 143
• 3. Survey in Korean 149
• 국문초록 156