Will intelligent upgrading lead to disorder? The impact of AI technology adoption on employees’ work

Abstract

Background

Artificial intelligence (AI) is a key driver of technological transformation. According to a McKinsey survey, 88% of organizations have already entered the nascent stage of AI technology. However, it is also changing workflows with implications for work–family balance.

Objectives

This study aims to investigate how AI technology adoption influences employees’ work–family conflict by analyzing the impact-response chain.

Methods

Grounded in conservation of resources theory, we surveyed 346 employees who had adopted AI technology at work. Data were analyzed using Mplus 8.3 and SPSS 27.0 to test the moderated sequential mediation model.

Results

AI technology adoption has both positive and negative effects on work–family conflict and is associated with increased conflict. However, job insecurity and leisure crafting mediate this relationship; job insecurity worsens conflict, whereas leisure crafting reduces conflict. Psychological detachment strengthens this effect by positively moderating the relationship.

Conclusion

AI’s impact on work–family conflict is not simply positive or negative, but both threat and opportunity, clarifying contradictions in past research and provides new answers for managing AI technology adoption.

Keywords

AI technology adoption job insecurity leisure crafting work–family conflict psychological detachment

Introduction

Artificial intelligence (AI), the theme of contemporary development,¹ has become a key driver of industrial transformation and economic growth. It has become a tool for organizational transformation, enabling cost reduction, efficiency gains, and sustainable competitive advantage. Consequently, its application and depth of integration continues to expand.² With the widespread adoption of AI, the technology is reshaping work processes and job content,³ posing real challenges for employees accustomed to traditional practices.⁴ Understanding employees’ AI technology adoption behavior is therefore critical for effective organizational management and digital transformation.

Previous studies on AI technology adoption primarily focused on antecedents from organizational, individual, and technical perspectives, including AI satisfaction, security, and privacy.^5,6 However, few studies have examined its outcomes, and those that do predominantly focus on one-sided effects, such as perceived technological threat or knowledge-sharing behavior.^7,8 Moreover, existing research has largely focused on the work domain, overlooking spillover effects into the family domain.⁹ This limits a full understanding of AI technology adoption, as work and family are interdependent life domains.¹⁰ We cannot reasonably understand the full effect of AI technology adoption on employees if we only consider its effect in the work domain and not its effect in the non-work domain.¹¹ Therefore, examining spillover effects beyond work is crucial for understanding AI’s implications for employee well-being during technological transition.¹² Earlier studies have insufficiently addressed subjective behaviors associated with AI technology adoption.¹³ Accordingly, this study focuses on the behavioral and psychological responses to AI technology adoption and its implications for employees’ work–family conflict.

AI technology adoption has altered work practices and benefit allocations,¹⁴ potentially increasing job insecurity.^15,16 Meanwhile, the emotions and behaviors experienced in the work domain can spill over into the family domain.¹⁷ As a typical workplace stressor, the negative effects of job insecurity might spill over into the non-work domain, leading to a major source of work–family conflict.¹⁸ Accordingly, job insecurity is modeled as a key mediator. By replacing repetitive and inefficient tasks, AI can free up employees’ time,¹⁹ enabling greater engagement in leisure crafting, which refers to individuals’ proactive attempts to improve their internal or external leisure environment for personal development and work-life balance.²⁰ Through leisure crafting, employees can arrange leisure activities in a more flexible and proactive manner, thereby obtaining psychological resources and developmental momentum.^21,22 This enables them to better pursue learning and development goals,²³ adapt to the occupational changes brought about by AI technology, and consequently bridge the tension between work and family roles, which will enable them to alleviate work–family conflict. Based on this, we propose another possible mediating variable and analyze the sequential mediating effects of job insecurity and leisure crafting. Gaining or losing resources is influenced by psychological traits. Employees with higher psychological detachment are more able to detach from their jobs physically and mentally,²⁴ and reinvest resources in leisure crafting because they are aware of resource investment. Thus, it can reduce work–family conflicts and help them access new resources. Therefore, psychological detachment was introduced as a moderating variable, and a moderated sequential mediation model was developed.

This study examines the spillover effect of employee AI technology adoption on the family domain and its impact mechanism on work–family conflict. This study also contributes to achieving work–family balance in management practices and extends prior research by linking technology, work, and family within an integrated framework. Moreover, by developing a moderated sequential mediation model with both positive and negative effects, it extends the conservation of resources theory to the digital transformation context.

Theoretical framework and development of hypotheses

Conservation of resources theory

This study applied the conservation of resources (COR) theory to examine how AI technology adoption affects employees’ work–family conflict. The COR theory, proposed by Hobfoll,²⁵ posits that individuals seek to obtain, retain, improve valued resources and that resource gains or losses directly influence psychological state. Resource loss at work generates stress and strain; however, individuals actively invest in and protect remaining resources to cope with actual or anticipated loss. Specifically, when employees adopt AI technology, they invest a great deal of resources in the work domain to deal with environmental pressures. If these resources are not adequately replenished, it will eventually lead to a reduction in resource investments in other domains. For an employee, the loss of resources in the family domain due to work makes it difficult for the person to meet their family responsibilities. However, to protect and acquire new resources, employees proactively engage in resource investments through various means to balance their resources. The COR theory has been applied across multiple domains, including workplace stress and work–family balance, providing a framework for understanding how individuals cope with stressful situations. This study explains the potential impact of AI technology adoption and highlights how employees can leverage resources to alleviate work–family conflict.²⁶ As AI becomes more embedded in the workplace, balancing one’s psychological resources and alleviating work–family conflicts thus becomes very important for the sustainable development of employees. This is also the core starting point for our exploration based on the COR theory.

AI technology adoption and work–family conflict

AI is any technology that correctly interprets external data and flexibly applies it to a specific task or goal.²⁷ Existing research on AI has primarily focused on the technology, including its adoption or acceptance.⁴ Individual AI technology adoption is the process by which an individual willingly accepts to use AI technology, integrating it into their tasks or routines. It encompasses an individual’s initial willingness or behavior to adopt,²⁸ as well as continued behavior after initial adoption, including the continuance, routinization, and adaptation of the technology.²⁹

COR theory posits that people tend to use key resources to cope with stressful situations. Resource scarcity increases self-protective responses, reducing resource investment in other domains.³⁰ AI technology adds higher job demands for employees. To adapt to technological change and overcome challenges posed by new work processes, employees may need to invest significant energy or time and are frequently confronted with technostress or emotional exhaustion arising from the complexities of the technology.³¹ If these depleted resources are not promptly replenished, an individual’s resources will continue to decline, triggering employees to engage in self-protection to avoid further resource loss. Work–family conflict is an inter-role conflict arising from incompatible work and family demands.³² If employees face continuous resource depletion from work demands without timely replenishment, they may reduce investments in the family domain. This prevents them from fulfilling their family responsibilities, leading to incompatibility between their work and family roles. Moreover, AI technologies are often characterized by their “anytime, anywhere” accessibility, which blurs the spatial and temporal boundaries between work and family, ultimately exacerbating the likelihood of work–family conflict.³³ Thus, we propose the following hypothesis:

AI technology adoption positively affects employee work–family conflict.

Mediating role of job insecurity

Job insecurity refers to employees’ perceived threat of losing their jobs and the associated uncertainty regarding continued employment. It involves worry and unease about losing their current job³⁴ and reflects a negative state arising from perceived inability to cope with change or uncertainty. With the rapid development and application of AI technologies, employees’ perceptions of job security are undergoing profound changes.¹⁵ Existing research confirms that the large-scale application of AI technology is closely related to job insecurity.³⁵ AI technology adoption causes employees to re-evaluate their own job value. This directly triggers survival anxiety about being replaced by AI technology, placing unprecedented pressure on employees’ job stability. According to COR theory, when employees struggle to gain new resources in the work domain, such as skill development or career advancement,^36,37 while simultaneously facing threats of resource loss, such as job status or income security,^38,39 they experience significant dual psychological pressure, which can manifest as job insecurity. As both an outcome of resource loss and a stressor, job insecurity further accelerates resource depletion,⁴⁰ triggering a loss spiral in which initial resource loss leads to further losses. Through role spillover, these effects extend from the work domain to the family domain.^41,42 Consequently, resource depletion undermines employees’ emotional regulation capacity as well as the time and energy to fulfill family responsibilities, thereby worsening work–family conflict. Thus, we propose the following hypotheses:

Job insecurity mediates the relationship between AI technology adoption and work–family conflict. AI technology adoption indirectly and positively affects work–family conflict through job insecurity.

Mediating role of leisure crafting

Leisure crafting refers to goal-oriented, proactive leisure activities that individuals pursue and implement to develop themselves, expand relationships, and enhance learning abilities.⁴³ It is an expansion of job crafting into the leisure domain, perceived as a special resource for escaping work pressure or emotional exhaustion. It emphasizes the dynamic behaviors in which individuals seek growth and development through leisure. Unlike the resource loss pathway, represented by job insecurity, leisure crafting reveals a more proactive resource investment pathway. Specifically, high job demands are more likely to trigger leisure-crafting behaviors in individuals.⁴⁴ The technological change and capability challenges brought about by AI represent a typical high job-demand status. Individuals may see this as a new impetus for development.⁴⁵ They may engage in crafting behaviors to compensate for potential negative reactions to work stress. This is an effective strategy for coping with sudden changes in the external environment. Therefore, leisure crafting holds significant resource potential, helping employees replenish the substantial energy expended at work,²¹ thereby enabling individuals to balance their work–family responsibilities and alleviate work–family conflict. According to the COR theory, driven by awareness of resource acquisition and investment, employees may engage in leisure crafting. This increases their resource-gaining behavior and actively builds and protects existing resources. Furthermore, leisure crafting can restore individual energy levels and offset energy depletion. Faced with the environmental uncertainty brought by AI technology adoption, employees achieve physical and mental relaxation through leisure activities.⁴⁶ Examples include participating in AI-themed puzzle games or using generative AI for creative projects. Through leisure activities, they can complete their resource replenishment and self-improvement. Existing research shows that online leisure crafting helps individuals achieve family thriving⁴⁷ and reduces conflicts. Leisure crafting can help individuals recover more quickly from resource depletion caused by AI technology adoption. Thus, we propose the following hypotheses:

Leisure crafting mediates the relationship between AI technology adoption and work–family conflict. AI technology adoption indirectly and negatively affects work–family conflict through leisure crafting.

Sequential mediating roles of job insecurity and leisure crafting

Individuals’ cognition is coherent, extending across work and non-work domains. Therefore, this study introduces two variables—job insecurity (work domain) and leisure crafting (non-work domain)—to examine their sequential mediating effects. During AI technology adoption, employees deplete their resources to cope with changes and uncertainty. This further increases job insecurity.⁴⁸ The initial loss of resources triggers further resource loss and causes job insecurity, as a stressor, to further deplete the employee’s personal resources. The COR theory posits that resource replenishment and gain are particularly important in the event of resource loss, especially for individuals with limited resources. In response, individuals build and protect their existing resource reserves. When employees cannot replenish their resource losses in the work domain, they proactively redirect their resource investments toward a more controllable leisure domain. In this context, leisure crafting shifts from passive recovery to an active strategy for resource replenishment. By setting and accomplishing meaningful leisure goals, employees rebuild the energy depleted by work within the non-work domain, thereby securing stable psychological resources outside of work. This enables them to return to their families with greater vitality, reducing the negative spillover from work to the family domain and consequently reducing work–family conflict. Therefore, leisure crafting becomes an effective strategy for coping with changes in the external environment.⁴⁴ Through this pathway, employees can protect their resources from depletion and thus better balance their work and family roles. Thus, we propose the following hypotheses:

Job insecurity and leisure crafting play sequential mediating roles in AI technology adoption and work–family conflict.

Moderating role of psychological detachment

Psychological detachment is defined as an individual’s ability to detach from work psychologically and physically during non-work time, without being preoccupied by work-related matters.⁴⁹ COR theory suggests that individual characteristics are likely to affect resource gain, loss, and depletion. Such characteristics affect individuals’ emotions, which in turn feed back into individual behavior through emotional reactions.⁵⁰ Existing research has shown that psychological detachment can serve as a recovery experience, helping employees replenish the energy consumed at work.⁵¹ In this context, psychological detachment significantly shapes whether the awareness of resource investment triggered by job insecurity can be effectively translated into actual resource gain. When employees feel insecure because of the impact of AI technology, those with high psychological detachment are better able to disengage from work-related strain, thereby reducing resource depletion. This enables them to redirect and invest resources in the leisure domain, where such investments can generate new resources and support personal development. Employees with low psychological detachment, even if driven by job insecurity, engage in resource investment behaviors (such as leisure crafting) to gain new resources, and their low level of detachment will prevent effective resource investment. Therefore, where job insecurity drives leisure crafting, employees with high psychological detachment can ensure that this resource investment—leisure crafting—produces expected resources and helps them achieve resource balance. Thus, we propose the following hypotheses:

Psychological detachment positively moderates the positive effect of job insecurity on leisure crafting.

Furthermore, this study considered a moderated sequential mediation model. Specifically, the stronger the psychological detachment, the more pronounced the positive effect of job insecurity on leisure crafting, thereby alleviating work–family conflict. This pathway represents a threat-coping mechanism. Job insecurity arising from the adoption of AI technology is typically perceived as a negative anticipation of resource loss. Paradoxically, when employees possess a high capacity for psychological detachment, this negative anticipation serves as a catalyst for engaging in leisure crafting. By mentally disengaging from work during non-work hours, individuals transform job insecurity into an opportunity for self-enhancement, facilitating resource recovery within the leisure domain, and ultimately preventing the negative spillover of insecurity into the family domain. Thus, psychological detachment not only helps employees engage in leisure crafting more effectively but also helps them balance role conflict between work and family to achieve resource balance. Eden⁵² resource replenishment during non-work hours leads to resource gain spirals. Whether from work breaks⁵³ or on weekends,⁵⁴ psychological detachment helps individuals detach themselves from work. This allows for successful resource recovery. For employees whose traditional work environment has changed owing to the adoption of AI technology, psychological detachment can help them learn and improve their skill levels during work breaks, to say nothing about vacations. This helps them better utilize their leisure time created by AI technology. This achieves a resource balance between work and family, reducing the dilemma of being unable to focus on family activities due to work pressures. The theoretical model is illustrated in Figure 1. In summary, we propose the following hypotheses:

Figure 1.

Theoretical model.

Psychological detachment moderates the sequential mediating effect of AI technology adoption on work–family conflict through job insecurity and leisure crafting.

Method

Sample and procedure

This study employed a cross-sectional survey design using standardized quantitative tools, consistent with established research practices.^55,56 Data were collected from December 2024 to January 2025 using a convenience sampling method to efficiently and rapidly access employees who frequently used AI technology in their work. This method is widely applied in organizational research as it balances feasibility and sample relevance.⁵⁷ Data collection was coordinated by a university-based survey research center with extensive experience in questionnaire administration. The center recruited and trained upper-level undergraduate or graduate students in survey procedures and ethical standards. After receiving instructions on the research objectives and sampling criteria, these assistants distributed the questionnaire through online channels, including personal contacts and social media platforms. All measurement scales were adapted from published literature in reputable journals, providing researchers with clear and reliable tools for investigation. The questionnaire comprised three parts. The first part introduced the study, assured anonymity and confidentiality, and provided completion instructions. Prior to the main survey, participants were asked whether they had used AI technology in their work. The study included only respondents who indicated adoption of AI technology in their work; those who did not meet this criterion were directed to exit the survey. The second part was the survey content, including AI technology adoption, job insecurity, leisure crafting, psychological detachment, and work–family conflict. The third part included demographic information, such as gender, age, education, and monthly income. To common method bias, procedural control measures recommended by Podsakoff et al.⁵⁸ were adopted, including anonymity assurance and reasonable setting of question order. A total of 442 questionnaires were distributed. After excluding invalid responses characterized by identical answers (e.g., selecting “5″ for all items) or patterned responses (e.g., a “1-2-3-4-5″ cycle), 346 valid questionnaires were obtained, yielding an effective response rate of 78.3%. The final valid sample consisted entirely of employed workers in China across 26 provinces, including Jilin, Shandong, and Guangdong. The sample was 40.5% male and 59.5% female; 56.6% were aged ≤25, 22.3% were 26–40, 17.9% were 41–55, and 3.2% were >55. Regarding education, 3.2% had junior high school or below, 7.8% high school/vocational, 16.8% college, 66.2% bachelor’s, and 6.1% master’s or above. Monthly income distribution was 50.9% below 4000 Yuan, 29.2% between 4001 and 8000 Yuan, 11.6% between 8001 and 12,000 Yuan, 3.2% between 12,001 and 16,000 Yuan, and 5.2% above 16,000 Yuan.

Measures

The questionnaire was fully examined, validated scales were employed to guarantee reliability, and a five-point Likert-type response format ranging from 1 (“strongly disagree”) to 5 (“strongly agree”) was used. AI technology adoption was evaluated through a set of eight items cited by Dong et al.,⁵⁹ with example items such as “I need AI to help me do my work” and a Cronbach’s alpha of 0.86. Job insecurity was evaluated through a set of five items cited by Mauno et al.,⁶⁰ with example items such as “Your job is insecure.” and Cronbach’s alpha of 0.758. Leisure crafting used the nine-item scale cited by Petrou and Bakker,⁴³ with example items such as “I try to build relationships through leisure activities,” and Cronbach’s alpha of 0.864. Work–family conflict employed the four-item scale cited by Liu et al.,⁶¹ with example items such as “After work, I often cannot relax due to work-related matters,” and Cronbach’s alpha of 0.748. Psychological detachment was assessed using a four-item scale cited by Sonnentag and Fritz,⁶² with example items such as “I forget about work.” and Cronbach’s alpha of 0.813. The Cronbach’s alpha for this study exceeded 0.7, which, according to the literature, indicates good reliability for all constructs.⁶³ In addition, drawing on previous studies, gender, age, education level, and monthly income were selected as control variables.

Results

Discriminant validity and confirmatory factor analysis

We ran a confirmatory factor analysis (CFA) on Mplus 8.3 to test the discriminant validity across variables of AI technology adoption, job insecurity, leisure crafting, work–family conflict, and psychological detachment. Following the recommendations of Schreiber et al.,⁶⁴ the indices used to assess the measurement model included the chi-square statistic (χ²), the root mean square error of approximation (RMSEA), the Tucker–Lewis index (TLI), and the comparative fit index (CFI). Table 1 shows that the five-factor model demonstrates a better fit than alternative models (χ² = 686.741, df = 395, χ²/df = 1.739, RMSEA = 0.046, SRMR = 0.051, CFI = 0.919, TLI = 0.911), confirming the distinctiveness of the five constructs and validating the good discriminant validity of the construct.

Table 1.

Results of confirmatory factor analysis.

Model	χ ²	df	χ²/df	RMSEA	SRMR	CFI	TLI
1. AIA,JI,LC,WFC,PD	686.741	395	1.739	0.046	0.051	0.919	0.911
2. AIA,JI+LC,WFC,PD	1045.949	399	2.621	0.068	0.081	0.821	0.805
3. AIA,JI,LC,WFC+PD	982.083	399	2.461	0.065	0.069	0.839	0.824
4. AIA,JI,LC+WFC+PD	1590.322	402	3.956	0.092	0.114	0.671	0.644
5. AIA,JI+LC+WFC+PD	1854.027	404	4.589	0.102	0.118	0.599	0.568
6. AIA +JI+LC+WFC+PD	2328.306	405	5.749	0.117	0.124	0.468	0.428

N = 346. AIA, AI technology adoption; JI, job insecurity; LC, leisure crafting; WFC, work–family conflict; PD, psychological detachment.

Common method variance

The independent, dependent, mediating, and moderating variables in this study were all self-reported; therefore, common method bias may exist. To address this, Harman’s single-factor test was conducted to examine common method variance. An unrotated exploratory factor analysis including all variables extracted six factors with eigenvalues greater than 1. The first factor accounted for 23% of the variance, which is below the threshold of 40 %, indicating no significant common method bias.

Descriptive statistics

Statistical descriptive analyses were performed using SPSS version 27.0. Table 2 shows the means, standard deviations, and correlation coefficients of all the variables. The results indicated that AI technology adoption is significantly positively correlated with job insecurity (r = 0.184, p < 0.01), leisure crafting (r = 0.432, p < 0.01), and work–family conflict (r = 0.138, p < 0.05). Job insecurity was significantly positively correlated with both work–family conflict (r = 0.321, p < 0.01) and leisure crafting (r = 0.332, p < 0.01). Leisure crafting was significantly and negatively correlated with work–family conflict (r = −0.111, p < 0.05). These findings align with theoretical expectations. Additionally, variance inflation factor (VIF) tests for all variables yielded values below 2, indicating little multicollinearity.

Table 2.

Descriptive statistics and correlations among study variables.

Variable	M	SD	1	2	3	4	5	6	7	8	9
1. Gender	0.595	0.492	1
2. Age	1.682	0.896	−0.142**	1
3. Edu	3.642	0.837	0.055	−0.268**	1
4. Income	1.827	1.092	−0.180**	0.423**	0.084	1
5. AIA	3.807	0.695	0.009	0.123*	−0.150**	−0.072	1
6. JI	3.528	0.763	0.050	0.062	−0.084	−0.052	0.184**	1
7. LC	3.790	0.629	−0.023	0.041	−0.111*	−0.049	0.432**	0.332**	1
8. WFC	3.382	0.776	−0.087	0.080	0.005	0.031	0.138*	0.321**	−0.111*	1
9. PD	3.473	0.906	0.023	0.186**	−0.075	−0.021	0.377**	0.307**	0.124*	0.254**	1

N = 346. AIA, AI technology adoption; JI, job insecurity; LC, leisure crafting; WFC, work–family conflict; PD, psychological detachment, the same below; *p < 0.05 **p < 0.01.

Hypothesis testing

Direct and mediation effect testing

Analyses were run using Model 6 in Process 4.2 macro. The analysis was conducted using the bootstrap method recommended by Preacher and Hayes.⁶⁵ This approach has become the prevailing paradigm for testing mediation effects and has been widely applied in empirical research across multiple disciplines, including psychology and sociology.⁶⁶ Compared with traditional methods, it does not rely on large sample sizes and effectively circumvents the potential biases inherent in conventional mediation effect estimation, thereby enhancing the reliability and replicability of the research findings. As such, it is particularly well-suited for this study. After standardizing all the variables, the hypothesized model was tested using 5000 bootstrap samples. Table 3 indicates that AI technology adoption significantly predicts work–family conflict (β = 0.213, p < 0.001), validating H1. AI technology adoption also significantly predicts job insecurity (β = 0.166, p < 0.01), and job insecurity significantly predicts work–family conflict (β = 0.401, p < 0.001). Further bootstrap analysis (Table 4) revealed that job insecurity mediates the relationship between AI technology adoption and work–family conflict, with an indirect effect of 0.067 and a 95% confidence interval of [0.019, 0.12], excluding zero. This confirms a significant indirect effect, supporting H2. AI technology adoption significantly predicts leisure crafting (β = 0.383, p < 0.001), and leisure crafting significantly predicts work–family conflict (β = −0.335, p < 0.001). The bootstrap test shows that leisure crafting mediates the relationship between AI technology adoption and work–family conflict, with an indirect effect of −0.128 and a 95% CI of [−0.192, −0.074], supporting H3. Furthermore, as presented in Table 3 and consistent with previous analysis, AI technology adoption positively affects job insecurity, and job insecurity positively influences leisure crafting (β = 0.263, p < 0.001), while leisure crafting negatively affects work–family conflict. Bootstrap results confirm that job insecurity and leisure crafting together form a sequential mediation pathway between AI technology adoption and work–family conflict. The sequential indirect effect was −0.015, with a 95% CI of [−0.031, −0.004], excluding zero, supporting H4. The total effect of AI technology adoption on work–family conflict is 0.137, with a 95% CI of [0.03, 0.244], excluding zero, indicating overall significance. This suggests that although AI technology adoption influences work–family conflict through both positive and negative pathways, it mainly exacerbates the conflict. The results of the path analysis integrating the sequential mediation model and moderation tests are summarized in Table 4 and illustrated in Figure 2.

Table 3.

Regression results for main, mediation effects.

Variable	Model 1: JI		Model 2: LC		Model 3: WFC
Variable	Value	SE	Value	SE	Value	SE
Gender	0.05	0.054	−0.042	0.048	−0.112^*	0.05
Age	0.06	0.063	−0.042	0.056	0.035	0.057
Income	−0.053	0.061	0.006	0.054	0.012	0.056
Edu	−0.042	0.057	−0.041	0.05	0.048	0.052
AIA	0.166^**	0.054	0.383^***	0.049	0.213^***	0.055
JI			0.263^***	0.048	0.401^***	0.052
LC					−0.335^***	0.056
R²	0.04		0.256		0.209
F	3.062*		19.466^***		12.736^***

∗p < 0.05. ∗∗p < 0.01. ∗∗∗p < 0.001, the same below.

Table 4.

Decomposition of direct and mediation effects.

Path	Coeff	BootSE	95% CI
Path	Coeff	BootSE	LLCI	ULCI
AIA→WFC	0.213	0.055	0.106	0.32
AIA→JI→WFC	0.067	0.026	0.019	0.12
AIA→LC→WFC	−0.128	0.031	−0.192	−0.074
AIA→JI→LC→WFC	−0.015	0.007	−0.031	−0.004

Figure 2.

Results of path coefficient analysis.

Moderation effect testing

After standardizing all the variables, Model 91 in Process 4.2 macro was employed for further analysis. The results, which are presented in Table 5, demonstrate that the interaction term between job insecurity and psychological detachment significantly and positively predicts leisure crafting (β = 0.173, p < 0.001), supporting Hypothesis 5. Simple slope analysis indicated that the positive effect of job insecurity on leisure crafting was more pronounced among employees with high psychological detachment. Accordingly, employees with greater psychological detachment were more inclined to invest resources in leisure crafting when driven by job insecurity (Figure 3), which further validates H5.

Table 5.

Analysis of moderating effects.

Variable	leisure crafting
Variable	Coeff	SE	t	95% CI
Gender	−0.035	0.046	−0.746	[−0.126, 0.057]
Age	−0.016	0.054	−0.291	[−0.123, 0.091]
Income	−0.003	0.052	−0.06	[−0.106, 0.1]
Edu	−0.009	0.05	−0.193	[−0.106, 0.087]
AIA	0.404^***	0.05	8.122	[0.306, 0.502]
JI	0.305^***	0.048	6.347	[0.21,0.399]
PD	−0.147^*	0.052	−2.853	[−0.249, −0.046]
JI × PD	0.173^***	0.039	4.448	[0.096, 0.249]
R²	0.31
F	18.801^***

Figure 3.

The moderating role of psychological detachment.

To further examine moderated mediation, the indirect effect of AI technology adoption on work–family conflict through job insecurity and leisure crafting was analyzed. For the high psychological detachment group, the indirect effect was −0.027, with a 95% CI of [−0.051, −0.008], excluding zero. In contrast, for the low psychological detachment group, the indirect effect was −0.007, with a 95% CI of [−0.022, 0.002], which includes zero. The difference between the indirect effects at high and low levels of psychological detachment was −0.019, with a 95% CI of [−0.039, −0.004], excluding zero. Thus, the moderated sequential mediation effect was statistically significant, confirming H6.

Discussion

Our study examined the effect of AI technology adoption on employees’ work–family conflict. Based on COR theory, we propose a moderated sequential mediation model. The model included job insecurity, leisure crafting, and psychological detachment as mediating variables. Our findings unpack multiple pathways and boundary conditions linking AI technology adoption to work–family conflict.

First, AI technology adoption exacerbates work–family conflict by increasing job insecurity. This aligns with prior research indicating that AI technology adoption results in negative effects,⁶⁷ such as job insecurity and a sense of technological threat,^35,68 which deplete employees’ psychological resources, thereby affecting their family life.⁶⁹ Based on this, this study verifies the indirect conflict that technological change imposes on employees’ non-work domains. Second, the study shows that AI technology adoption can reduce work–family conflict via a leisure crafting pathway. This pathway indicates that employees are not passive when faced with technological change. Related research has indicated that AI technology can lead employees to form a challenge appraisal, which in turn fosters positive employee behaviors or strengthens behavioral motivation,⁵⁹ thereby facilitating outcomes such as job crafting.⁷⁰ This study reveals that under the influence of AI technology, employees are able to proactively adjust their leisure activities to achieve a better balance between work and family life, supporting the view that, in some circumstances, high job demands can lead to active coping strategies used by employees⁷¹ and breaks the singular vision of employees as passive pressure recipients of technological change. Third, driven by job insecurity, employees use leisure crafting intentionally to reduce the risk of resource loss and acquire new resources, which allows them to coordinate their work and family roles, reducing work–family conflict. This demonstrates the compensatory and spillover effects of leisure crafting and offers a new perspective for interpreting the internal logic of employees’ stress and adaptation after AI technology adoption. This finding further illustrates that contexts of resource loss may activate individuals’ proactive resource investment awareness to forestall further resource depletion, which is consistent with the assertion made by Hobfoll et al.³⁰ regarding the driving mechanism of resource investment. Fourth, individuals with higher psychological detachment performed better at transforming job insecurity into motivation. They participate in active leisure crafting to cope with stress, and work–family conflict is less likely. As a psychological resource, psychological detachment enhances an individual’s adaptability and resilience in the context of technological change, reflecting their capacity to shift attention from resource threats at work to resource investment in the leisure domain. This finding aligns with the perspective of Sonnentag et al.⁵¹ that psychological detachment facilitates employee recovery and energy replenishment. Additionally, AI technology adoption positively and negatively affects work–family conflict. This differs from the one-sided view of previous research on AI technology adoption that AI effects are either positive or negative. Adopting a resource gain-loss perspective, this study examines how AI technology adoption reshapes work–family conflict. Prior research has largely portrayed employees as passive recipients of technological change, emphasizing resource depletion, diminished well-being,⁷² and heightened work–family conflict. Although this perspective identifies potential threats, it overlooks employee agency and the capacity for resource investment under stressful circumstances. AI technology entails both “threats and opportunities.”⁷³ Consistent with this view, prior research reveals that AI technology can stimulate employees’ developmental motivation.⁵⁹ Building on this premise, the present study identifies a previously underexplored compensatory pathway—namely, the buffering effect of leisure crafting. The findings indicate that AI technology adoption does not merely intensify conflict through a resource depletion pathway; rather, AI technology and the job insecurity it engenders can prompt proactive self-enhancement in the leisure domain. Through leisure crafting, employees replenish resources and develop capabilities, thereby mitigating work–family conflict. This discovery extends prior research, which has predominantly focused on negative effects. This suggests that AI-induced strain does not inevitably intensify conflict in a unidirectional manner. Instead, the impact of AI technology adoption emerges from the concurrent operation of both resource depletion and resource investment mechanisms. Therefore, this provides an integrated framework for fully understanding the complex effects of technological change on employees’ work–family relationships.

Theoretical implications

This study enriches the literature on the outcomes of AI technology adoption. The adoption of AI technology in the workplace may have multiple impacts on employees’ psychology and behavior,¹³ which may spill over to various aspects of life and family. By introducing job insecurity and leisure crafting as mediating variables, this study develops an integrated “technology-work-family” analytical framework and extends AI technology adoption outcomes from the work domain to the family domain, linking work, leisure and family systems. This approach enhances understanding of non-work outcomes and clarifies multiple pathways through which AI technology adoption affects work–family conflict, thereby providing a more comprehensive perspective for future research. Second, this study addresses the limitations of one-dimensional perspectives on AI technology adoption outcomes by examining its complex transmission mechanisms. Prior research has indicated that when confronted with emerging technologies, employees exhibit distinct psychological reactions and behaviors contingent upon variations in individual characteristics and environmental factors,^74,75 and these effects are often neither positive nor negative. Grounded in COR theory, this study develops a moderated sequential mediation model capturing both resource loss and gain processes. It thus reveals the multiple “threat-compensation” pathways of AI technology adoption: on one hand, it increases work–family conflict by leading to job insecurity; on the other hand, it reduces conflict via leisure crafting, which enables resource compensation and spills over into the family domain. In addition, this study clarifies that if individuals encounter job insecurity (perceived resource threat) as a result of AI technology adoption, they are not completely put into passive defense. Instead, driven by awareness of resource acquisition, they may proactively engage in resource investment and reconstruction through leisure crafting. Their goals were to compensate for resource loss in the work domain and prevent further depletion. This finding clearly presents the logic behind employees’ proactive adjustments to technological changes. It provides an integrated explanatory framework to resolve contradictions in past research that offer singular positive or negative conclusions. Finally, this study extends COR theory to digital transformation contexts by explaining how resource loss and recovery processes operate in an AI-enabled work environment. It demonstrates how work-domain resource dynamics spill over into the family domain, thereby broadening the theory’s applicability and explanatory power in contemporary organizational settings.

Management insights

This study primarily focuses on the multiple pathways through which AI technology adoption affects employees’ work–family conflict and the moderating role of psychological detachment. Enterprises should pay more attention to the multilayer impact of AI technology on employees’ psychology and behavior in the digital-intelligent transformation. Managers should implement systemic management measures to help employees achieve a balance between work and family life.

First, organizations should anticipate and intervene in the pressure that AI technology adoption places on employees. When confronted with new technologies, employees inevitably experience a degree of apprehension and psychological resistance.⁷⁶ The research confirms that AI technology adoption adversely influences work–family conflict in a direct way or does so indirectly by increasing job insecurity. Organizations can implement regular AI technology training and adaptation courses and clarify that AI technology is an assistive rather than replacement technology, helping employees understand that technological change may shape job value but does not replace roles. At the same time, it is important to ensure transparent employee communication channels. Timely communication of a company’s AI strategy and development plans can reduce employee anxiety about job continuity. This addresses the source, reducing resource depletion caused by job insecurity and its spillover effects into the family domain. Second, organizations may encourage employees to engage in leisure crafting and achieve a high degree of psychological detachment. Our research shows that leisure crafting can effectively resolve work–family conflict. Thus, employees with higher psychological detachment can better translate job insecurity into motivation for resource investment. Well-utilized leisure time can enhance employees’ sense of meaning and self-efficacy, thereby facilitating comprehensive self-improvement.²¹ Organizations should respect and protect employees’ non-work time. They need to ensure that after completing work tasks, employees can enjoy both fragmented and extended leisure time for their leisure activities. Furthermore, organizations can help finance leisure activities to gain skills and cultivate interests. This guides employees to transform their leisure time into opportunities for resource regeneration and capability enhancement, thereby shifting from passive consumption to active construction. Finally, build a “technology-work-family” support system holistically. The essence of organizational competition lies fundamentally in the contest for talent and the safeguarding of employee well-being.⁷⁷ Integrating employee well-being into both organizational development strategies and the evaluative criteria for digital and intelligent transformation is conducive to achieving sustainable organizational development. Employees’ work–family status and AI technology usage are regularly checked through surveys or interviews. Establish Employee Assistance Programs (EAP) to provide psychological counseling and resource referral services for employees and their families facing significant transformation pressure. Employees’ family roles are considered in management decision-making. For example, when adopting or assigning AI systems, think about how they affect the employee’s schedule and family responsibilities. Make appropriate humanized adjustments to this effect so that the organization can respect and support employees’ full-life roles.

Limitations and prospects

Our study has some limitations. First, the data were self-reported, which may have introduced common method bias. Future research should utilize multi-time, multi-channel, and repeated measure designs to mitigate this issue. Second, although this study controlled for demographic variables such as sex, age, education level, and monthly income, other confounding factors may still exist. For instance, organizational-level variables (e.g., perceived organizational support, organizational culture, leadership style) may influence employees’ AI technology adoption^78,79; family-level variables (e.g., presence of minor children, spouse employment status, family caregiving responsibilities) may have a direct impact on work–family conflict.^80,81 Additionally, individual variables (e.g., personality traits, technology self-efficacy) may play a role in the relational pathways examined in this study.^82,83 Furthermore, we did not collect information on organizational characteristics (e.g., industry, size), limiting further analysis. Future research should employ multilevel modeling to incorporate macro-level variables (e.g., organization type and depth of AI technology adoption) and better identify boundary conditions of AI effects on work–family relationships. Finally, the cross-sectional design limits causal inference. As AI technology adoption is dynamic, future research should use longitudinal designs to track organizational AI implementation and examine changes in employees’ psychological and behavioral responses over time.

Footnotes

ORCID iD

Yanhong Guo

Ethical considerations

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent from the patients/participants or patients/participants legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

Author contributions

YG: Conceptualization, Project administration, Resources, Supervision, Validation, Writing—review & editing. LW: Methodology, Formal analysis, Software, Writing—original draft, Writing—review & editing.

Funding

The authors declare that financial support was received for the research and/or publication of this article. The present research was supported by the Youth Fund Project of Humanities and Social Sciences Research, Ministry of Education, 2025 (Project Title: The Impact of AI Technology Adoption on Two-Way Intergenerational Knowledge Transfer among the “Cross-Generational Labor Force” and Its Acceleration Mechanism, Grant No: 25YJC630035).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.*

References

Wang

Xing

Zhang

. AI technology application and employee responsibility. Hum Soc Sci Commun 2023; 10: 356. https://doi.org/10.1057/s41599-023-01843-3

Enholm

Papagiannidis

Mikalef

, et al. Artificial intelligence and business value: a literature review. Inf Syst Front 2022; 24: 1709–1734. https://doi.org/10.1007/s10796-021-10186-w

Man Tang

Koopman

McClean

, et al. When conscientious employees meet intelligent machines: an integrative approach inspired by complementarity theory and role theory. Acad Manag J 2022; 65: 1019–1054. https://doi.org/10.5465/amj.2020.1516

Cheng

Lin

Kong

. Challenge or hindrance? How and when organizational artificial intelligence adoption influences employee job crafting. J Bus Res 2023; 164: 113987. https://doi.org/10.1016/j.jbusres.2023.113987

Dennis

Lakhiwal

Sachdeva

. AI agents as team members: effects on satisfaction, conflict, trustworthiness, and willingness to work with. J Manag Inf Syst 2023; 40: 307–337. https://doi.org/10.1080/07421222.2023.2196773

Lee

Emokpae

, et al. What makes you continuously use chatbot services? Evidence from Chinese online travel agencies. Electron Mark 2021; 31: 575–599. https://doi.org/10.1007/s12525-020-00454-z

Gao

Agafari

, et al. How and when artificial intelligence adoption promotes employee knowledge sharing? The role of paradoxical leadership and technophilia. Front Psychol 2025; 16: 1573587. https://doi.org/10.3389/fpsyg.2025.1573587

Zhou

Chen

. A mixed-methods investigation of the factors affecting the use of facial recognition as a threatening AI application. Internet Res 2024; 34: 1872–1897. https://doi.org/10.1108/intr-11-2022-0894

Jeong

. Organizational AI investment and its ripple effects: an investigation of job insecurity, coworker conflict and work–family conflict. Pers Rev 2025; 54: 1729–1751. https://doi.org/10.1108/PR-06-2024-0578

10.

Peng

Chen

, et al. The relationship between work–family conflict and health behaviors: a systematic review and meta-analysis. Behav Sci 2026; 16: 386. https://doi.org/10.3390/bs16030386

11.

Deng

Ding

Chen

. Techno-stressors, work–family conflict and task performance: exploring the dual moderating effects of family and intrinsic motivation. Ind Manag Data Syst. 2026; 1–23. https://doi.org/10.1108/IMDS-12-2024-1281

12.

Zhou

Rasiah

, et al. An empirical study on the dark side of service employees’ AI awareness: behavioral responses, emotional mechanisms, and mitigating factors. J Retailing Consum Serv. 2024; 79: 103869. https://doi.org/10.1016/j.jretconser.2024.103869

13.

Jia

Ning

Liu

. The consequences and theoretical explanation of workplace AI on employees: a systematic literature review. Journal of Digital Management 2025; 1: 14. https://doi.org/10.1007/s44362-025-00016-3

14.

Orabi

Alahewat

AMM

Abualfalayeh

, et al. The interdisciplinary nature of ai and human resource management: a bibliometric analysis. Hum Syst Manag. 2024; 43: 845–871. https://doi.org/10.3233/hsm-240054

15.

T-J

Liang

Wang

. The buffering role of workplace mindfulness: how job insecurity of human-artificial intelligence collaboration impacts employees’ work–life-related outcomes. J Bus Psychol 2024; 39: 1395–1411. https://doi.org/10.1007/s10869-024-09963-6

16.

Kim

B-J

Kim

M-J

. How artificial intelligence-induced job insecurity shapes knowledge dynamics: the mitigating role of artificial intelligence self-efficacy. J Innov Knowl 2024; 9: 100590. https://doi.org/10.1016/j.jik.2024.100590

17.

Pluut

Ilies

, et al. How social stressors at work influence marital behaviors at home: an interpersonal model of work–family spillover. J Occup Health Psychol 2022; 27: 74–88. https://doi.org/10.1037/ocp0000298

18.

Sora

Caballer

Peiró

. Spillover outcomes of job insecurity: differences among men and women (Consecuencias indirectas de la inseguridad laboral: diferencias entre hombres y mujeres). International Journal of Social Psychology 2021; 36: 149–178. https://doi.org/10.1080/02134748.2020.1840233

19.

Ertel

Bonenberger

. Rebound effects caused by artificial intelligence and automation in private life and industry. Sustainability 2025; 17: 1988. https://doi.org/10.3390/su17051988

20.

Han

M-C

Hwang

P-C

. Crafting job and leisure activities when you are overqualified. J Hospit Tourism Manag 2021; 48: 146–154. https://doi.org/10.1016/j.jhtm.2021.05.017

21.

Petrou

de Vries

. Context-free and work-related benefits of a leisure crafting intervention: a randomized controlled trial. J Leisure Res 2025; 56: 123–148. https://doi.org/10.1080/00222216.2023.2244953

22.

Petrou

Hamrick

Abdel

. Unraveling the power of leisure crafting for unengaged employees: implications for creativity and meaning at work. J Organ Behav 2024; 45: 1170–1188. https://doi.org/10.1002/job.2824

23.

Chen

I-S

Choi

. Does leisure crafting improve workplace creativity? Eur Manag Rev 2025; 22: 640–655. https://doi.org/10.1111/emre.12669

24.

Agolli

Holtz

. Facilitating detachment from work: a systematic review, evidence-based recommendations, and guide for future research. J Occup Health Psychol 2023; 28: 129–159. https://doi.org/10.1037/ocp0000353

25.

Hobfoll

. Conservation of resources: a new attempt at conceptualizing stress. Am Psychol 1989; 44: 513–524. https://doi.org/10.1037/0003-066X.44.3.513

26.

Yan

Zhang

, et al. Conflict or enrichment? A comprehensive review of workfamily balance. Int J Conflict Manag 2025; 37: 221–251. https://doi.org/10.1108/ijcma-05-2025-0142

27.

Raisch

Krakowski

. Artificial intelligence and management: the automation–augmentation paradox. Acad Manag Rev 2021; 46: 192–210. https://doi.org/10.5465/amr.2018.0072

28.

Khasawneh

. Concepts and measurements of innovativeness: the case of information and communication technologies. Int J Arab Cult Manag Sustain Dev 2008; 1: 23–33. https://doi.org/10.1504/IJACMSD.2008.020487

29.

Limayem

Hirt

Cheung

. How habit limits the predictive power of intention: the case of information systems continuance. MIS Q 2007; 31: 705–737. https://doi.org/10.2307/25148817

30.

Hobfoll

Halbesleben

Neveu

J-P

, et al. Conservation of resources in the organizational context: the reality of resources and their consequences. Ann Rev Organ Psychol Organ Behav 2018; 5: 103–128. https://doi.org/10.1146/annurev-orgpsych-032117-104640

31.

Chen

Zhao

Wang

. The effect of job skill demands under artificial intelligence embeddedness on employees’ job performance: a moderated double-edged sword model. Behav Sci 2024; 14: 974. https://doi.org/10.3390/bs14100974

32.

Greenhaus

Beutell

. Sources of conflict between work and family roles. Acad Manag Rev 1985; 10: 76–88. https://doi.org/10.5465/amr.1985.4277352

33.

Regina

Allen

. Taking rivalries home: workplace rivalry and work-to-family conflict. J Vocat Behav 2023; 141: 103844. https://doi.org/10.1016/j.jvb.2023.103844

34.

Sverke

Hellgren

. The nature of job insecurity: understanding employment uncertainty on the brink of a new millennium. Appl Psychol 2002; 51: 23–42. https://doi.org/10.1111/1464-0597.0077z

35.

Koo

Curtis

Ryan

. Examining the impact of artificial intelligence on hotel employees through job insecurity perspectives. Int J Hospit Manag 2021; 95: 102763. https://doi.org/10.1016/j.ijhm.2020.102763

36.

Bao

Liu

, et al. When job resources function as a stress buffer: a resource allocation perspective of the job demands-resources model. Pers Indiv Differ 2022; 192: 111591. https://doi.org/10.1016/j.paid.2022.111591

37.

Arranz Lahuerta

López Ramajo

Gandía

. Artificial intelligence and organisational transformation: technical skills, job insecurity and adoption. Manag Decis. 2026; 1–16. https://doi.org/10.1108/MD-05-2025-1462

38.

Allen

Hammer

Mohr

, et al. Conserving what’s left: a longitudinal investigation of how daily financial worry fuels perceived stress and drives turnover intentions and job search behaviors. J Bus Psychol. 2025; 1–21. https://doi.org/10.1007/s10869-025-10083-y

39.

Son

Yang

Park

. How organizational politics and subjective social status moderate job insecurity–silence relationships. J Manag Organ 2023; 29: 266–286. https://doi.org/10.1017/jmo.2022.54

40.

Yang

Zhao

Zhu

, et al. How daily job insecurity links to next-day ingratiation: the roles of emotional exhaustion and power distance orientation. Psychol Res Behav Manag 2024; 17: 2807–2818. https://doi.org/10.2147/PRBM.S438242

41.

Zou

H-Y

Wang

H-J

, et al. Qualitative job insecurity, negative work-related affect and work-to-family conflict: the moderating role of core self-evaluations. J Career Dev 2023; 50: 216–233. https://doi.org/10.1177/08948453221090610

42.

Petitta

Probst

Ghezzi

, et al. Job insecurity and work–family interface as predictors of mental and physical health: the moderating role of family–work stereotype threat. J Occup Organ Psychol 2024; 97: 452–478. https://doi.org/10.1111/joop.12478

43.

Petrou

Bakker

. Crafting one’s leisure time in response to high job strain. Hum Relat 2016; 69: 507–529. https://doi.org/10.1177/0018726715590453

44.

Petrou

Bakker

Van Den Heuvel

. Weekly job crafting and leisure crafting: implications for meaning‐making and work engagement. J Occup Organ Psychol 2017; 90: 129–152. https://doi.org/10.1111/joop.12160

45.

Cai

Zhao

. How AI-related task complexity shapes innovative work behavior: a coping theory perspective. Behav Sci 2025; 15: 1467. https://doi.org/10.3390/bs15111467

46.

Abdel

Bakker

Häusser

. The role of leisure crafting for emotional exhaustion in telework during the COVID-19 pandemic. Anxiety Stress Coping 2021; 34: 530–544. https://doi.org/10.1080/10615806.2021.1903447

47.

Chen

I-S

. Turning home boredom during the outbreak of COVID-19 into thriving at home and career self-management: the role of online leisure crafting. Int J Contemp Hospit Manag 2020; 32: 3645–3663. https://doi.org/10.1108/IJCHM-06-2020-0580

48.

Kim

B-J

Lee

. AI adoption, employee depression and knowledge: how corporate social responsibility buffers psychological impact. J Innov Knowl 2025; 10: 100815. https://doi.org/10.1016/j.jik.2025.100815

49.

Chong

Kim

Lee

, et al. Mind your own break! the interactive effect of workday respite activities and mindfulness on employee outcomes via affective linkages. Organ Behav Hum Decis Process 2020; 159: 64–77. https://doi.org/10.1016/j.obhdp.2019.11.001

50.

Weiss

Cropanzano

. Affective events theory: a theoretical discussion of the structure, causes and consequences of affective experiences at work. Res Organ Behav 1996; 18: 1–74.

51.

Sonnentag

Kark

Venz

. Leader support for recovery: a multi‐level approach to employee psychological detachment from work. J Occup Organ Psychol 2024; 97: 1762–1788. https://doi.org/10.1111/joop.12538

52.

Eden

. Job stress and respite relief: overcoming high-tech tethers. In: Perrewe

Ganster

(eds). Exploring Theoretical Mechanisms and Perspectives. Emerald Group Publishing Limited, 2001, pp. 143–194.

53.

Coffeng

van Sluijs

Hendriksen

, et al. Physical activity and relaxation during and after work are independently associated with the need for recovery. J Phys Activ Health 2015; 12: 109–115. https://doi.org/10.1123/jpah.2012-0452

54.

Fritz

Sonnentag

Spector

, et al. The weekend matters: relationships between stress recovery and affective experiences. J Organ Behav 2010; 31: 1137–1162. https://doi.org/10.1002/job.672

55.

Lin

Tian

Cheng

. Facilitation or hindrance: the contingent effect of organizational artificial intelligence adoption on proactive career behavior. Comput Hum Behav 2024; 152: 108092. https://doi.org/10.1016/j.chb.2023.108092

56.

Maier

Thatcher

Grover

, et al. Cross-sectional research: a critical perspective, use cases, and recommendations for IS research. Int J Inf Manag 2023; 70: 102625. https://doi.org/10.1016/j.ijinfomgt.2023.102625

57.

Zickar

Keith

. Innovations in sampling: improving the appropriateness and quality of samples in organizational research. Ann Rev Organ Psychol Organ Behav 2023; 10: 315–337. https://doi.org/10.1146/annurev-orgpsych-120920-052946

58.

Podsakoff

MacKenzie

Lee

J-Y

, et al. Common method biases in behavioral research: a critical review of the literature and recommended remedies. J Appl Psychol 2003; 88: 879–903. https://doi.org/10.1037/0021-9010.88.5.879

59.

Dong

Tian

, et al. When knowledge workers meet AI? The double-edged sword effects of AI adoption on innovative work behavior. J Knowl Manag 2025; 29: 113–147. https://doi.org/10.1108/jkm-02-2024-0222

60.

Mauno

Leskinen

Kinnunen

. Multi‐wave, multi‐variable models of job insecurity: applying different scales in studying the stability of job insecurity. J Organ Behav: The International Journal of Industrial, Occupational and Organizational Psychology and Behavior 2001; 22: 919–937. https://doi.org/10.1002/job.122

61.

Liu

Zhou

Shi

. Research on negative outcomes of organizational citizenship behavior: based on generalized exchange,impression management and evolutionary psychology(in Chinese). Manag Rev 2017; 29: 163–180. https://doi.org/10.14120/j.cnki.cn11-5057/f.2017.04.016

62.

Sonnentag

Fritz

. The recovery experience questionnaire: development and validation of a measure for assessing recuperation and unwinding from work. J Occup Health Psychol 2007; 12: 204–221. https://doi.org/10.1037/1076-8998.12.3.204

63.

Tavakol

Dennick

. Making sense of Cronbach's alpha. Int J Med Educ 2011; 2: 53–55. https://doi.org/10.5116/ijme.4dfb.8dfd

64.

Schreiber

Nora

Stage

, et al. Reporting structural equation modeling and confirmatory factor analysis results: a review. J Educ Res 2006; 99: 323–338. https://doi.org/10.3200/JOER.99.6.323-338

65.

Preacher

Hayes

. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods 2008; 40: 879–891. https://doi.org/10.3758/BRM.40.3.879

66.

MacKinnon

Lockwood

Williams

. Confidence limits for the indirect effect: distribution of the product and resampling methods. Multivariate Behav Res 2004; 39: 99–128. https://doi.org/10.1207/s15327906mbr3901_4

67.

Xue

. Unemployment risk perception and knowledge hiding under the disruption of artificial intelligence transformation. Soc Behav Personal Int J 2023; 51: 1–12. https://doi.org/10.2224/sbp.12106

68.

Kamboj

. Mapping the complex interactions between work-family conflict, AI technostress, work-life balance and employees’ well-being: moderated mediation analysis. Int J Conflict Manag. 2026; 37: 713–741. https://doi.org/10.1108/IJCMA-08-2025-0274

69.

Zheng

Zhang

. Association between AI awareness and emotional exhaustion: the serial mediation of job insecurity and work interference with family. Behav Sci 2025; 15: 401. https://doi.org/10.3390/bs15040401

70.

Wang

, et al. Opportunity and threat: how employees’ perceptions of artificial intelligence influence job crafting. Baltic J Manag 2025; 20: 511–527. https://doi.org/10.1108/BJM-09-2024-0525

71.

Tadić

Bakker

Oerlemans

. Challenge versus hindrance job demands and well‐being: a diary study on the moderating role of job resources. J Occup Organ Psychol 2015; 88: 702–725. https://doi.org/10.1111/joop.12094

72.

Pan

. How and when organizational artificial intelligence adoption impacts employees' well-being. Int J Ment Health Promot 2025; 27: 1769–1780. https://doi.org/10.32604/ijmhp.2025.070147

73.

Chuang

Y-T

Chiang

H-L

Lin

A-P

. Insights from the job demands–resources model: AI's dual impact on employees’ work and life well-being. Int J Inf Manag 2025; 83: 102887. https://doi.org/10.1016/j.ijinfomgt.2025.102887

74.

Castaneda

Maseeh

Surachartkumtonkun

, et al. Impact of frontline employees’ perceived benefits of artificial intelligence on AI adoption and job engagement: a meta-analysis. Int J Hospit Manag 2026; 133: 104446. https://doi.org/10.1016/j.ijhm.2025.104446

75.

Verma

Singh

Tudoran

, et al. Elevating employees’ psychological responses and task performance through responsible artificial intelligence. Inf Technol People 2024; 37: 2551–2567. https://doi.org/10.1108/ITP-05-2023-0431

76.

Golgeci

Ritala

Arslan

, et al. Confronting and alleviating AI resistance in the workplace: an integrative review and a process framework. Hum Resour Manag Rev 2025; 35: 101075. https://doi.org/10.1016/j.hrmr.2024.101075

77.

Brammer

Millington

Rayton

. The contribution of corporate social responsibility to organizational commitment. Int J Hum Resour Manag 2007; 18: 1701–1719. https://doi.org/10.1080/09585190701570866

78.

Divya

Jain

Chetty

, et al. The mediating effect of leadership in artificial intelligence success for employee-engagement. Manag Decis 2025; 63: 3676–3699. https://doi.org/10.1108/md-01-2024-0213

79.

Lopes

Gomes

Nogueira

, et al. Exploring psychosocial predictors of AI use intention: the moderating role of organisational culture in diverse workplaces. Behav Inf Technol. 2026; 1–19. https://doi.org/10.1080/0144929X.2026.2626743

80.

Stas

Vanroelen

, et al. Linked lives and work-family dynamics: gendered workload, family support, and mental well-being among urban Chinese dual-earner households. Int J Equity Health 2025; 24: 312. https://doi.org/10.1186/s12939-025-02687-w

81.

Schnettler

Concha-Salgado

Orellana

, et al. Workload, job, and family satisfaction in dual-earning parents with adolescents: the mediating role of work-to-family conflict. Front Psychol 2025; 16: 1529092. https://doi.org/10.3389/fpsyg.2025.1529092

82.

Leong

AMW

Bai

Rasheed

, et al. AI disruption threat and employee outcomes: role of technology insecurity, thriving at work, and trait self-esteem. Int J Hospit Manag 2025; 126: 104064. https://doi.org/10.1016/j.ijhm.2024.104064

83.

Yin

Jiang

Niu

. Can AI really help? The double-edged sword effect of AI assistant on employees’ innovation behavior. Comput Hum Behav 2024; 150: 107987. https://doi.org/10.1016/j.chb.2023.107987

Will intelligent upgrading lead to disorder? The impact of AI technology adoption on employees’ work–family conflict

Abstract

Background

Objectives

Methods

Results

Conclusion

Keywords

Introduction

Theoretical framework and development of hypotheses

Conservation of resources theory

AI technology adoption and work–family conflict

Mediating role of job insecurity

Mediating role of leisure crafting

Sequential mediating roles of job insecurity and leisure crafting

Moderating role of psychological detachment

Method

Sample and procedure

Measures

Results

Discriminant validity and confirmatory factor analysis

Common method variance

Descriptive statistics

Hypothesis testing

Direct and mediation effect testing

Moderation effect testing

Discussion

Theoretical implications

Management insights

Limitations and prospects

Footnotes

ORCID iD

Ethical considerations

Author contributions

Funding

Declaration of conflicting interests

Data Availability Statement

References