Artificial Intelligence Symbolic Leadership in Small and Medium-Sized Enterprises: Enhancing Employee Flexibility and Technology Adoption

Abstract

This study examines the influence of leaders’ artificial intelligence symbolization on job-crafting behaviors, highlighting both positive and negative consequences in Chinese small and medium-sized firms. This research utilizes signaling theory to investigate the impact of leaders’ visible adoption of AI on employees’ readiness for change, perceived threats, and job-crafting behaviors. This study examines the moderating influence of organizational support to understand its amplifying and decreasing effects. This work utilizes Python-based statistical tools to provide a novel approach for evaluating behavioral data in social science research. The results reveal that leaders’ AI symbolization significantly improves employees’ readiness for change and promotes proactive job crafting. Conversely, symbolic actions may exacerbate perceived risks, adversely affecting job-crafting behaviors. Organizational support is essential to enhancing the beneficial impacts of AI symbolization on change readiness while alleviating its adverse consequences on perceived threats. These results show how crucial symbolic leadership is for getting people to use new technology and making staff more flexible in SMEs that use AI. By offering organizational training and resources, leaders may optimize favorable results and mitigate adverse effects. This study highlights its significance regarding change readiness, perceived threats, and job crafting. Furthermore, it underscores Python’s (3.9) potential as a groundbreaking tool for enhancing behavioral research in the age of AI.

1. Introduction

Artificial intelligence (AI) is an intricate technology developed to simulate human intelligence and improve performance [1]. Several analysts believe that AI will significantly enhance the quality of life for most people in the next ten years [2]. As we enter the era of artificial intelligence, its impact will be seen globally and in almost every industry. Algorithms, robotics, and machine learning are becoming essential in several areas due to the rapid development of AI [3]. Chinese small and medium-sized enterprises (SMEs) that focus on project-based work increasingly use AI to improve their ability to compete and stimulate innovation. These organizations possess notable flexibility and agility, making them suitable for implementing AI-driven solutions. The discourse surrounding AI adoption in SMEs is prevalent; nonetheless, a significant gap exists in comprehending how leaders’ AI symbolization, characterized by their proactive endorsement and advocacy of AI, affects employee outcomes like change readiness, perceived threat, and job crafting. The current literature has predominantly concentrated on the technical and operational dimensions of AI implementation while neglecting the symbolic actions of leaders and their psychological impact on employees. This study investigates the dual function of leaders’ AI symbolization in promoting employee adaptation and alleviating perceived threats, specifically within the framework of Chinese SMEs.

As AI becomes prevalent in organizations, leaders increasingly demonstrate actions that symbolize AI in the workplace. The leaders in these SMEs are rapidly adopting AI to revamp conventional workflows and obsolete technology to enhance efficiency and foster creativity. This perspective is consistent with previous studies on symbolization, which demonstrated how an organization’s people communicate their ideals and personal attributes to their colleagues in the workplace [4]. Jack Ma, Alibaba’s CEO, is enthusiastic about integrating AI features into the company’s most recent interior designs to demonstrate its commitment to innovation [5]. Integrating AI into the workplace substantially impacts leaders’ preferences and values [6]. Therefore, examining the effects of AI symbolization by leaders on employee job results is essential, a subject that has not been thoroughly researched. Similar to other proactive leadership styles like transformational leadership [7] and ethical leadership [8], AI symbolized by leaders is crucial to influencing employee views and behaviors. Recent research indicates that symbolic leadership practices might substantially affect employees’ readiness to accept change [9]. The precise effect of AI symbolization on employee outcomes is inadequately investigated, especially inside SMEs, where resource limitations and organizational support systems significantly differ from those in international enterprises.

Leaders’ AI symbolization, like other proactive and negative behaviors such as job crafting, taking charge, innovative behavior, stress and anxiety, resistance to change, and employee perceived threat, is characterized by a forward-looking and transformational approach [10,11]. By implementing AI-driven projects, leaders demonstrate the significance of adopting cutting-edge technologies to improve organizational performance and competitiveness. This symbolization acts as a catalyst for change, motivating staff to embrace new AI tools and approaches. Nevertheless, implementing AI in the workplace might pose difficulties, leading to heightened tension, anxiety, and reluctance to adapt to new circumstances. This is due to people becoming concerned about the potential consequences of AI on their job positions. Understanding the dual nature of AI symbolization by leaders is essential for effectively managing its impact on employee behavior and organizational results. As AI becomes more prevalent in the workplace, there is a growing trend of leaders adopting behaviors that embrace AI-driven changes. However, the effects of these leadership behaviors on employee outcomes have not been thoroughly investigated. Moreover, the extent to which organizational support contributes to resolving these challenges has not been extensively investigated. SMEs provide a unique setting for examining AI adoption owing to their agility, adaptability, and limited resources [12]. In contrast to more prominent companies, SMEs frequently lack institutionalized support networks, rendering employees more vulnerable to perceived threats and resistant to change [13]. Despite the increasing prevalence of AI in SMEs, research on the impact of leaders’ symbolic behavior on employee outcomes in this environment remains scarce. This study addresses this gap by investigating the relationship among leaders’ AI symbolization, organizational support, and employee responses in Chinese SMEs.

This gap exists due to the recent recognition by academics in organizational behavior of the significant impact of AI on organizational management [3,6,14]. However, some studies have examined how leaders’ AI-related actions impact employee outcomes, particularly regarding job crafting [15]. The influence of organizational support in this area has not been extensively studied; more specifically, there has been a lack of comprehensive research on the impact of organizational support on employee change readiness and perceived threats. This lack of attention is especially pertinent in SMEs, where resources and support systems may vary substantially from larger organizations. This research is based on signaling theory [16], which asserts that leaders’ behaviors and actions operate as signals that shape employee perceptions and behaviors. In AI adoption, leaders’ representation of AI conveys a message to employees on the significance and worth of AI technology. Recent research has utilized signaling theory to elucidate the influence of leadership behaviors on employee attitudes regarding organizational change [17]. We assert that leaders’ representation of AI conveys the advantages of AI adoption, hence improving employees’ readiness for change and diminishing their perceived threat.

To address this issue, we propose that leaders’ AI symbolization, which involves leaders actively demonstrating and promoting the usage of AI, could influence employees’ perceptions of threat. The utilization of AI can cause employees to experience anxiety or perceive it as a threat, perhaps resulting in unfavorable outcomes. Comprehending these dynamics has substantial implications on employee job performance, which is defined as the ability to fulfill job duties and accomplish essential tasks [15]. Our approach suggests that how leaders represent AI directly impacts how employees shape their job responsibilities. This influence occurs through two parallel mediators: the employees’ readiness to adapt to change and their perception of the potential threats AI poses. Employee change readiness refers to the extent to which employees are willing to adapt to new procedures driven by AI. In contrast, employees’ perceived threat refers to the worry and fear that employees experience concerning the application of AI. Moreover, we propose that the influence of leader AI symbolization on both mediators is influenced by organizational support. Providing resources and encouragement for adopting AI can improve employees’ readiness for change and reduce their perception of threats. Examining these relationships can substantially contribute to the expanding literature on human–AI interactions in organizational management and provide insights for leadership development in the age of artificial intelligence [1,6,15]. As AI technologies become more common in many industries, leaders can use this knowledge to enhance staff performance and job crafting while decreasing perceived risks and unethical conduct in SMEs. This research offers multiple contributions to the body of literature regarding AI adoption and leadership. Initially, it expands signaling theory by contextualizing it within the framework of AI symbolization in SMEs. Secondly, it examines the deficiency in comprehension regarding the impact of leaders’ symbolic actions on employee outcomes, especially in resource-limited settings. Ultimately, it offers pragmatic insights for leaders and companies on efficiently managing AI implementation’s psychological and behavioral implications.

This study utilizes signaling theory [16] to assess the possible effects of leaders’ AI symbolization on favorable staff outcomes. Signaling theory posits that acts and behaviors indicate personal intents and preferences [18]. Contextual signals can impact people’s perception of a situation and subsequent behaviors, as demonstrated by [19]. This study proposes that when leaders use AI symbolization, it stimulates employees’ change readiness. Change readiness refers to employees’ emotional and cognitive preparedness to accept and embrace changes. This, in turn, motivates employees to proactively make changes in their work, such as job crafting. Moreover, this study suggests that leaders’ use of AI symbolization also prompts employees’ perception of potential dangers, leading to anxiety and panic related to adopting AI technology. Implementing AI technology may cause employees to experience feelings of job insecurity and uncertainty about their future employment. This imagined threat can have a detrimental effect on their job-crafting behaviors. Leaders need to support, influence, and regulate these relationships. Providing tools, training, and encouragement for adopting AI can enhance the relationship between leaders’ representation of AI and employees’ readiness for change. When employees see substantial organizational support, they are more inclined to adopt AI and actively participate in proactive job design. On the other hand, support from the company could weaken the relationship between leaders’ representation of artificial intelligence and employees’ perception of threat. Providing sufficient assistance to employees reduces the likelihood of them perceiving AI as a threat, allowing them to concentrate on utilizing AI to improve their work.

Understanding the dynamics of Chinese SMEs, in which resources and support systems may vary dramatically from larger organizations, is essential. Leaders’ representation of AI can be a powerful signal to employees, highlighting the significance of embracing AI technologies. Insufficient organizational support might cause employees to interpret these signals as threats, resulting in feelings of worry and reluctance toward change. Chinese SMEs can boost employees’ willingness to adopt AI technology and encourage proactive job-crafting behaviors by implementing strong support systems, ultimately reducing perceived threats. This study examines the impact of leaders’ AI symbolization on staff job-crafting behaviors in Chinese SMEs. This research examines the positive and negative ramifications of AI adoption by leaders, concentrating on how these symbolic acts influence employees’ readiness for change, perceived threats, and their propensity for job crafting. This research is based on signaling theory, which posits that leaders’ behaviors and signals affect employee perceptions and actions. This research investigates the moderating influence of organizational support on the impact of AI symbolization on employee outcomes, determining whether support systems enhance or mitigate these effects. This study seeks to provide significant insights into leadership approaches within the AI era, specifically in SMEs with distinct resource allocation issues.

This paper is structured as follows: the Introduction provides the research background, identifies gaps in existing literature, and outlines the study’s objectives. The Literature Review and Hypothesis Development Section highlights relevant studies on AI adoption, leadership, and job crafting within the under-explored domain of artificial intelligence symbolism. The Methodology Section elaborates extensively on data analysis and research strategy. The Results and Discussion underscore the significance of artificial intelligence representation by leaders and the mediating role of organizational support, revealing critical findings. Theoretical and Practical Implications emphasize how much the research enhances academic theory and leadership style. The Conclusion and Future Directions summarize the findings and suggestions for future study areas.

2. Theoretical Background and Hypotheses

Signaling theory is a significant concept that originated from Spence’s work in 1973 within labor markets [18]. This concept elucidates how one entity (the “signaler”) transmits information to another entity (the “receiver”) via signals that mitigate information asymmetry. The theory posits that signals (actions, symbols, or behaviors) can mitigate uncertainty in social or corporate transactions by providing stakeholders with information regarding unobservable attributes, such as intentions or capabilities. Signaling theory has developed over time and is now extensively utilized in fields such as management, organizational behavior, and marketing, where leaders and organizations deploy symbols and behaviors to convey intentions to employees or external stakeholders [16]. Within the realm of AI integration in SMEs, signaling theory offers a comprehensive framework for comprehending how leaders’ AI symbolization, characterized by their proactive display and acceptance of AI, affects employee perceptions and behaviors. The symbolic representation of AI by leaders conveys the significance and value of AI technology, influencing employees’ perceptions of change and their readiness to adopt AI-driven innovations. Recent studies have utilized signaling theory to examine the impact of leadership behaviors on staff outcomes during technological transitions [17], yet the precise function of AI symbolization is inadequately investigated, especially in SMEs.

Signaling theory is based on two assumptions: signaling reduces information asymmetry (when one side has more information) and signal reliability (since misleading signals may be punished, signal dependability and how well they represent reality determine their value). Within the framework of Chinese SMEs and AI symbolization, leaders’ AI symbolization is a signal that mitigates employees’ confusion concerning the influence of AI on their jobs. Leaders who favorably represent AI may articulate its potential advantages, such as innovation and increased productivity, thereby shaping employees’ attitudes and behaviors. This is especially pertinent regarding Chinese SMEs, where resource limitations and informal organizational frameworks enhance the significance of leaders’ symbolic actions in influencing employee reactions to AI implementation [12]. In Chinese SMEs, adopting and integrating AI, essential for innovation and technological adaptation, necessitates effective communication to address employees’ concerns regarding AI-related changes. Leaders indicate the implications of AI on their roles through their actions and symbols. This influences employee behavior, including job crafting, which refers to the proactive alterations employees implement in their work tasks, relationships, or cognitive approaches to their roles [20]. Symbolic representation of AI by leaders can yield conflicting consequences on employee outcomes. Positive indicators regarding AI adoption can improve employees’ readiness for change and encourage proactive job crafting. Conversely, confusing or negative signals may intensify employees’ perceived threats, resulting in resistance and anxiousness. This duality highlights the need to comprehend how leaders’ AI representation affects employee conduct within SMEs, where the implications of AI implementation are more critical due to constrained resources and support systems.

Signaling theory offers a framework for examining the impact of leaders’ AI symbolization on employees’ attitudes and behaviors, encompassing both positive and negative effects. There are several inconsistencies and contradictions in the literature concerning the integration of AI in SMEs, especially in Chinese contexts, despite the substantial use of signaling theory in other situations. Although signaling theory has been extensively utilized in organizational behavior and leadership research, its applicability to AI representation in SMEs is still restricted. The current study has predominantly concentrated on the technical and operational dimensions of AI implementation, neglecting the symbolic actions of leaders and their psychological impact on employees. This study investigates how leaders’ AI symbolization affects employee change preparedness, perceived threat, and job crafting in Chinese SMEs, offering a unique viewpoint on the significance of signaling in AI-driven organizational transformation.

Within the framework of signaling theory, leaders’ representation of AI conveys significant signals to employees, influencing their perceptions and reactions to its use in the workplace. Leaders who favorably represent AI might encourage employees to perceive this technology as an opportunity instead of an imminent threat. It is crucial to influence employees’ engagement with their work, prompting them to modify or tailor their activities, relationships, and methodologies to fit AI integration better. When employees see leaders’ AI signals favorably, they are more inclined to accept change and demonstrate a willingness to adapt. Organizational support is essential to enhancing the beneficial impacts of leaders’ AI representation. Organizations can bolster leaders’ signals by offering resources, training, and encouragement, thus improving employees’ preparedness for change and diminishing their perceived dangers. This is especially crucial in SMEs, where formal support systems may be deficient, and staff depend significantly on leadership cues to manage AI-driven transformations [12].

Unambiguous and positive indications from leadership diminish uncertainty, emphasizing the advantages of AI, including enhanced job performance, which cultivates confidence and readiness to undertake additional responsibilities. When leaders convey confusing or negative signals, employees may perceive AI as a threat, fearing job insecurity or diminished significance in their positions. This perceived threat may elevate resistance to change and obstruct adaptation efforts. Organizational support plays a crucial role in this context. An encouraging environment bolsters affirmative signals from leadership, assisting employees in perceiving AI as a valuable asset and alleviating apprehensions regarding possible risks. Robust organizational support increases the likelihood that employees would view AI as an opportunity, enhancing engagement and adaptability in their tasks. This assistance can help alleviate the impact of ambiguous signals, ensuring that staff remain receptive to AI-driven transformations and enhancements. Signaling theory elucidates how leadership, AI representation, and organizational support influence employee behavior and preparedness for change through these linkages.

2.1. The Effect of Leaders’ AI Symbolization

Signaling theory illuminates how leaders’ AI symbolization affects employees’ perceptions and behaviors. Initially formulated by Spence (1973) for labor markets, signaling theory posits that signals such as leadership behavior communicate essential information that influences recipients’ perceptions and decisions. In AI, leaders’ symbolic communication functions as a strategic instrument to alleviate uncertainty and synchronize staff expectations [21]. Leaders’ use of symbols to convey AI’s significance and value affects employees’ job crafting, change readiness, and risk perceptions. Leaders can influence employee behavior by explicitly demonstrating AI’s potential benefits, lowering technological transition uncertainty. Precise and positive signals are vital in workplaces undergoing transitions due to technological progress. Positive leadership signals about AI have improved employee outcomes, including job crafting and change readiness. Leaders identifying AI as a growth and innovation opportunity might inspire workers to job craft or rework their positions to accommodate technology. Some researchers have shown that the effects of leadership signals differ by organization [2,3]. Due to resource and organizational constraints, SMEs struggle to reduce employee threat perceptions [22]. When leaders’ AI symbolization is vague or inconsistent, employees’ perceptions and responses to technological developments vary, resulting in contradictory results. In Chinese SMEs, leaders’ AI symbolization and employee responses are crucial. These companies must incorporate new technology to be competitive, but staff may not have much experience with AI. Effective leaders encourage staff to reduce anxieties and encourage positive engagement with AI. Leadership signaling affects SMEs’ ability to reconcile technical innovation with employee well-being and job happiness, making it essential to investigate [23]. Organizational growth depends on understanding these dynamics, especially when people feel vulnerable to new technologies. Based on these theoretical foundations, we propose three direct hypotheses.

Hypothesis H1.

Leaders’ AI symbolization positively influences employee job crafting.

Research indicates that when leaders effectively represent AI, people are more inclined to engage in job crafting to align their responsibilities with technology advancements. Leaders who highlight the advantages of technological integration can inspire staff to transform their work environments to actively enhance innovation [24]. Empirical research indicates that leadership communication significantly influences work-crafting behaviors. Research by Wrzesniewski and Dutton [20] indicates that when employees perceive support from leadership, they actively adjust their responsibilities and work identity to conform to organizational objectives. In the realm of AI adoption, leaders’ affirmative framing motivates people to perceive AI as an enabler instead of a disruptor [25].

Hypothesis H2.

Leaders’ AI symbolization positively influences employee change readiness.

Leaders’ proactive representation of AI increases employees’ willingness to accept change. Empirical evidence indicates that the manner in which leadership frames technological change directly influences workforce adaptability. Research by Rafferty and Griffin [26] demonstrates that when leaders present AI as an opportunity instead of a disruption, people display enhanced psychological safety and receptiveness to change. A recent study on digital transformation indicates that leader-driven AI narratives mitigate opposition by promoting a culture focused on learning [27]. Leaders who present AI as a creative and advantageous instrument mitigate uncertainty and foster trust, prompting people to be more adaptive and receptive to workplace transformations [28].

Hypothesis H3.

Leaders’ AI symbolization negatively influences employees’ perceived threat.

Conversely, when leaders convey ambiguous or adverse signals regarding AI, employees are more inclined to view AI as a job security risk, heightening anxiety and resistance to change. This applies to Chinese SMEs, where employees may exhibit heightened sensitivity to employment structure alterations due to economic uncertainty [29]. In collectivist work cultures such as China, concerns regarding job security substantially affect employees’ reactions to the adoption of AI [30]. Recent research [2,31] indicates that unclear AI signals from leadership may increase employee anxiety, especially in SMEs, where job restructuring is viewed as a heightened threat. The findings highlight the necessity of clear leadership communication to reduce uncertainty.

2.2. The Effect of Change Readiness and Perceived Threat

Change preparedness and perceived threat influence employee engagement with new technology, including AI. The Technology Acceptance Model (TAM) [32] holds that employees’ impressions of usefulness and simplicity of use determine their degree of readiness for technological change. Psychological safety and leadership communication help further shape change readiness [33]. On the other hand, the Job Demands–Resources (JD-R) model [34] contends that perceived work insecurity brought on by artificial intelligence functions as a psychological demand, which can cause resistance and disengagement. Studies indicate that employees with greater change readiness are more inclined to modify their roles and responsibilities to correspond with technology progress, participating in job crafting. Change readiness denotes employees’ psychological and behavioral preparedness to accept change, promoting proactive participation in their work [11,35]. Conversely, perceived threats, mainly work security or role significance, adversely impact job crafting. When employees perceive AI as a potential risk, it results in resistance and disengagement, obstructing proactive job crafting [29]. Numerous research studies have investigated the correlation between change readiness and perceived threats in technological transition and employee behavior [3,15]. Suseno et al. [36] discovered that change readiness substantially improves employees’ capacity to adjust to organizational changes by promoting proactive activities such as job designing. Tims et al. [37] discovered that perceived concerns diminish employees’ interest in job crafting, particularly concerning job security. This research underscores a dual dynamic in which preparation for positive change promotes adaptability, whereas perceived risks engender resistance. Numerous research studies on this theme consistently demonstrate that change preparation positively impacts job crafting, but perceived threats negatively influence it [36,37]. The extent of influence may differ depending on the organizational situation. Zhou and Velamuri [24] contend that in SMEs, the influence of perceived threats may be amplified due to the restricted resources available to mitigate the effects of technological progress. Conversely, in larger firms with more robust support systems, the adverse impacts of perceived threats may be diminished. Examining the correlation among leaders’ AI representation, change readiness, perceived risks, and job crafting is especially significant within the framework of Chinese SMEs. SMEs in China frequently encounter resource limitations and experience pressure to implement new technology, such as artificial intelligence, to sustain competitiveness. Employees in these firms may have restricted familiarity with AI, rendering their adaptability to change and their view of AI a dangerous critical factor for successful technology integration [38]. Empirical research on Chinese SMEs reveals that employees often feel a dual reaction to AI adoption: enthusiasm for innovation and fear of redundancy [39]. Given the collectivist culture and hierarchical work context, leadership messaging is vital in shaping views of technological change [30]. When leaders explicitly position AI as enablers rather than disruptors, people are likelier to adopt a proactive posture toward job crafting [40]. Comprehending these dynamics enables firms to develop superior leadership communication methods and support systems that alleviate dangers and promote adaptability. Leaders’ symbolic representation of AI influences job crafting via two parallel pathways: enhancing change readiness and perhaps engendering perceived risks. Through signaling theory, leaders’ AI representation serves as a signal that can enhance favorable outcomes while also intensifying adverse impressions. Leaders who explicitly communicate the advantages of AI promote readiness for change, facilitating job crafting as people feel equipped to interact with new technology. Conversely, confusing or negative signals can evoke perceived dangers, hindering employees’ inclination to customize their roles. Based on these theoretical foundations, we propose two indirect hypotheses.

Hypothesis H4.

Leaders’ AI symbolization increases employees’ change readiness, enhancing job crafting.

Research indicates that when leaders effectively communicate the beneficial effects of AI, people are more inclined to accept the changes and actively participate in job crafting [37]. Research on digital transformation suggests that leadership framing strongly impacts employees’ change readiness [41]. A study by Rafferty and Griffin [26] indicated that effective leadership communication minimizes uncertainty and promotes psychological safety, making employees more inclined to engage in proactive job designing. In AI integration situations, leaders who prioritize growth and learning opportunities enable people to synchronize their activities with new technologies [25].

Hypothesis H5.

Leaders’ AI symbolization may also lead to perceived threats, negatively affecting job crafting.

Research indicates that when employees interpret confusing or negative signals regarding AI, they perceive it as a threat, resulting in disengagement and resistance to job crafting [42]. Studies in organizational psychology reveal that perceived job uncertainty causes psychological disengagement tendencies [26]. High ambiguity about the use of AI among employees increases their likelihood of disengaging out of concern about redundancy [43]. In SMEs, this effect is stronger, as limited retraining prospects intensify concerns about job loss [24].

2.3. The Effect of Organizational Support

Organizational support significantly influences employees’ perceptions and reactions to leadership cues, especially concerning AI representation. Within the framework of technological transition, organizational support denotes the resources, training, and encouragement provided to employees by their employer to facilitate adaptation to new technology. Research indicates that robust organizational support can enhance positive leadership signals, bolstering employees’ preparedness for change and diminishing the perceived risks of AI implementation [44]. The Perceived Organizational Support (POS) theory [45] proposes that employees view organizational support as an indication that their contributions are valued and their well-being is prioritized. Within AI implementation, POS increases the legitimacy of leadership signals, lowering ambiguity and encouraging psychological safety [46]. This coincides with signaling theory, where employees’ views of organizational backing impact their responses to leaderships’ AI communication. Prior studies have repeatedly shown that organizational support enhances the correlation between leadership behaviors and employee outcomes [47]. Gigliotti et al. [48] discovered that individuals who recognize substantial organizational support are more inclined to react favorably to leadership initiatives, exhibiting enhanced readiness for change and engagement with new technology. Furthermore, research indicates that organizational support can mitigate adverse impacts. Zhang et al. [49] demonstrated that organizational support alleviates perceived dangers and diminishes resistance to change by cultivating a sense of security among employees. This indicates that organizational support amplifies the beneficial effects of leadership signals while mitigating their adverse consequences. Most research indicates that organizational support consistently amplifies the benefits of leadership signals, including heightened preparedness for change and job crafting. Nonetheless, there exist specific contextual variances. Langer et al. [50] propose that in smaller enterprises, such as SMEs, the accessibility of organizational resources may be constrained, diminishing organizational assistance’s efficacy in mitigating perceived threats. Larger firms may provide more extensive support networks, hence enhancing the beneficial impacts of leadership signaling. Comprehending the significance of organizational assistance is especially crucial in Chinese SMEs, where employees frequently encounter difficulties associated with resource limitations and technological unpredictability. Studies on Chinese SMEs [51] suggest that insufficient financial and training resources intensify employees’ perceived threats toward AI adoption. Unlike larger firms with well-funded reskilling programs, SMEs typically struggle to provide formal support systems, making leadership communication even more crucial in molding employees’ technology perspectives [24]. This underlines the necessity for precise organizational assistance to combat perceived threats. In some firms, the mere representation of AI leadership may be inadequate to promote favorable employee results. Robust organizational support is crucial for assisting employees in adapting to the changes introduced by AI implementation, as it can improve their preparedness for transformation and mitigate perceived risks. This is especially significant in SMEs, where the workforce may have limited familiarity with modern technologies such as AI, rendering organizational assistance crucial for successful technology integration [52]. Signaling theory offers a comprehensive framework for comprehending the interplay between organizational support and leadership signals. Leaders who positively represent AI convey messages that affect employees’ willingness to accept change. Organizational support can enhance these signals, magnifying the beneficial effects of leadership on employee readiness for change. Conversely, when leaders’ AI representation is unclear or unfavorable, organizational support can serve as a buffer, mitigating the perceived threat by equipping staff with the necessary resources and assurances to feel comfortable amid technological transitions. Based on these theoretical foundations, we propose two moderating and two moderated mediation hypotheses.

Hypothesis H6.

Organizational support strengthens the positive relationship between leaders’ AI symbolization and employee change readiness.

Studies demonstrate that when employees recognize substantial organizational support, the beneficial impact of leaders’ AI symbolization on change preparedness is enhanced. Employees who perceive organizational support are more inclined to adopt new technologies and proactively adjust their responsibilities to align with technological progress [53]. Empirical research supports this moderating impact. Studies by Rafferty and Minbashian [54] indicated that employees in firms with excellent support systems exhibit improved change readiness due to less transition anxiety. Furthermore, a meta-analysis by Whelan-Berry and Somerville [55] reveals that organizational support boosts employees’ ability to absorb and implement leadership-driven technical initiatives, including AI adoption.

Hypothesis H7.

Organizational support weakens the negative relationship between leaders’ AI symbolization and employees’ perceived threat.

Research indicates that organizational support mitigates the adverse impacts of ambiguous or unfavorable AI representation by leaders. When employees see support, their view of AI as a threat diminishes, allowing for more positive engagement with AI [52]. This is consistent with studies in occupational stress theory [56], which propose that during organizational transformation, strong support structures serve as a coping mechanism. By increasing employees’ perceived influence over technology transitions, Aldabbas et al. [57] found that organized training programs and career development activities help to reduce AI-induced anxiety.

Hypothesis H8.

Organizational support strengthens the indirect effect of leaders’ AI symbolization on job crafting through employee change readiness.

Leaders’ symbolic use of AI affects employees’ job-crafting behaviors by improving their readiness for change. This link is considerably enhanced by strong organizational support. Employees who recognize substantial organizational support are more inclined to take leadership cues favorably and demonstrate a willingness to adapt, resulting in heightened job-crafting behaviors. Based on studies in Job Design Theory [20], this point of view is confirmed by implying that workers who have access to developmental resources voluntarily redesign their jobs to fit technological developments. Furthermore, a longitudinal study by Tims et al. [58] showed that workers in companies with excellent technology and managerial support systems show more job-creating activities, especially when the leadership cues are clear and encouraging. Prior studies substantiate this connection, demonstrating that organizational support improves change preparedness and promotes a proactive stance toward role adaptation and innovation [59]. Consequently, organizational support enhances the mediating function of change preparedness in the association between leaders’ AI symbolization and job crafting.

Hypothesis H9.

Organizational support weakens the indirect effect of leaders’ AI symbolization on job crafting through employee perceived threat.

Conversely, leaders’ AI symbolism may adversely impact job crafting because of perceived threats, mainly when the signals conveyed by leaders are vague or unfavorable. Nevertheless, robust organizational support can mitigate the adverse effects of perceived threats. Employees who see organizational support are less inclined to regard AI as a danger, especially without explicit information from leadership regarding AI. Studies indicate that organizational support alleviates the fear and anxiety linked to technological change, diminishing job-crafting resistance [36]. This is consistent with psychological safety theory [60], which holds that workers who feel comfortable in their workplace are more ready to try out new roles and technology. Stankevičiūtė et al. [61] conducted a study showing that organized HR interventions, including AI upskilling programs, lower employees’ uncertainty about technological redundancy, thus raising their willingness to engage in job creation. Organizational support mitigates the adverse mediating effect of perceived risks on work crafting. The conceptual model of the study is presented in Figure 1.

3. Methodology

An online questionnaire survey was conducted utilizing convenience sampling to evaluate our theoretical model within the context of Chinese SMEs. The survey collected data from employees in SMEs that utilize AI technology. This approach facilitated examining the relationships among leaders’ AI symbolization, employee change readiness, perceived threat, and job crafting. The online survey method facilitated efficient data collection, yielding insights into real-world organizational dynamics while preserving external validity [62]. Data were collected from two technology sectors in China that feature AI: healthcare and e-commerce/retail. High-tech companies focus on researching, developing, and producing AI-powered healthcare solutions, including medical imaging and diagnostics. AI-driven e-commerce platforms use AI for personalized recommendations and dynamic inventory management. The emphasis of these companies on AI research and development corresponds effectively with our theoretical framework. The data collected from these two sectors enabled the generation of a sample representative of our target population, comprising employees in AI-driven technological environments. A total of 376 full-time employees were invited to participate in this study. Surveys were distributed through WeChat, a popular messaging application with over one billion active users [63]. All participants provided informed consent and engaged voluntarily without receiving any compensation. Identification codes were employed to match participants’ responses throughout the survey, thereby ensuring confidentiality. The initial survey was created in English and translated into Chinese to guarantee clarity and cultural relevance. Based on Brislin’s [64] recommendations for cross-cultural research, we employed the back-translation method to ensure accuracy. The English questionnaire was translated into Chinese by a bilingual expert. This method confirmed the validity of the translated questionnaire, rendering it appropriate for the target population. The collected data were coded and analyzed in Chinese before being translated into English for reporting and analysis. A bilingual expert translated Chinese into English to maintain the original meaning.

The collected information was processed utilizing Python, a multifaceted programming language extensively employed for statistical analysis and structural equation modeling (SEM). Python was chosen for its powerful modules, including semopy and statsmodels, which enable the estimation and assessment of intricate SEM models. These methods allowed us to analyze the interconnections between leaders’ AI symbolization, employee change preparedness, perceived threat, and job crafting with accuracy and adaptability. The SEM analysis was performed in Python by following these steps: Data Preparation: The dataset underwent a cleaning process to address missing values and outliers. The data underwent standardization and were formatted into a covariance matrix to meet the requirements of structural equation modeling (SEM). Model Specification: The SEM framework was established utilizing the semopy library. Latent variables, including leaders’ AI symbolization and employee change readiness, were defined alongside their relationships to observed variables. The structural and measurement models were coded in Python with explicit detail. Model Estimation: The estimation of the model was conducted utilizing Maximum Likelihood Estimation (MLE), which was executed through the semopy library. This method was selected due to its robustness and appropriateness for complex models. Model Evaluation: The assessment of model fit was conducted utilizing various fit indices, which include SRMR (Standardized Root Mean Square Residual), d_ULS (Unweighted Least Squares Discrepancy), d_G (Geodesic Discrepancy), Chi-square, and NFI (Normed Fit Index). The selected indices assessed the model’s goodness-of-fit and confirmed its correspondence with the data. Results Interpretation: The final parameter estimates and significance levels were extracted and analyzed to assess the hypothesized relationships among the constructs. The model fitness indices validated the robustness of the SEM model, thereby confirming the validity of the findings. During the SEM analysis, a challenge was identified due to missing data. This issue was resolved through the application of multiple imputation techniques. Model convergence issues were addressed through the refinement of the model specification and the verification that the data adhered to the assumptions of SEM.

Each measurement scale was mapped to the corresponding latent variables in the SEM model, as described below. All measurements in this study were obtained from previous research with confirmed reliability and validity. All scale items were evaluated using a five-point Likert scale, from 1 (strongly disagree) to 5 (strongly agree). The readiness for employee change was assessed with a nine-item scale from Rafferty and Minbashian [54]. The intention for employee job crafting was evaluated using a six-item measure developed by Chen et al. [40]. The AI symbolization behaviors of leaders were assessed using a nine-item scale derived from the moral symbolization scale by Desai and Kouchaki [19]. Organizational support was assessed utilizing a six-item scale developed by Hutchison [65]. Perceived threat was evaluated utilizing a six-item measure developed by Ethier and Deaux [66].

3.1. Data Analysis

Python has become essential in social sciences, especially business studies [67]. It facilitates practical data analysis and the implementation of advanced machine-learning methodologies. Its adaptability and availability make it a preferred option for researchers and practitioners pursuing actionable insights from complex datasets. Ref. [68] examined Python’s application in decision sciences, providing real-world case studies demonstrating its efficacy across numerous industries, including social sciences and business research. The first step in our data analysis included setting up the environment by loading the necessary Python plugins. Pandas was used for effective data manipulation and analysis. Semopy was used to perform SEM to analyze the interrelations among leaders’ AI symbolization, employee change readiness, perceived threat, and job crafting. Upon configuring the libraries, the dataset was imported into a Pandas Data Frame via the pd.read_csv() function, with the file path indicating the data file’s location. The data.head() command was employed to verify the correct loading of the dataset and to comprehend its structure by presenting the initial rows, hence providing a swift overview of the column names and data content. This fundamental phase established the basis for all ensuing data investigations and analyses. The next step of our data analysis involved comprehensively examining the dataset to identify its structure and verify its integrity. This comprehensive examination guaranteed that the dataset was prepared for analysis, facilitating precise and reliable outcomes in the following stages.

Upon completing the preliminary data inspection and verifying the dataset’s accuracy and completeness, we reestablished the environment for the demographic analysis. This entailed re-importing the necessary libraries. After examining the dataset and verifying its integrity and completeness, we used Python packages to assess its demographic distribution. The dataset classified healthcare and e-commerce workers into three employment groups: software developers, data analysts, and researchers. We generated industry employee distributions and examined job types within each industry. The dataset included 376 employees from these sectors and employment categories. Healthcare employed 53.2%, or 200 people, and e-commerce employed 46.8%, or 176 people. Software developers made up 39.9% (150 employees), data analysts 31.9% (120 individuals), and researchers 28.2% (106 employees). We noticed that there were 80 software developers, 70 data analysts, and 50 researchers in healthcare using group-level aggregation. E-commerce employed 70 software developers, 50 data analysts, and 56 researchers, emphasizing their technical and analytical roles.

3.2. Measurement Model

The study measured five primary constructs: LAS, ECR, EPT, OSA, and EJC. For each construct, multiple items examined factor loadings, Cronbach’s Alpha, composite reliability (CR), average variance extracted (AVE), and variance inflation factors. These measures ensure construct reliability, validity, and noncollinearity. All the detailed values related to reliability and validity, including factor loadings, Cronbach’s Alpha, composite reliability (CR), average variance extracted (AVE), and variance inflation factors (VIF), are presented in

Table 1. Validity and reliability analysis.

Constructs	Items	Factor Loadings	Cronbach Alpha	CR	AVE	VIF
Leaders’ AI Symbolization	LAS 1	0.633	0.939	0.939	0.632	1.710
	LAS 2	0.747				1.811
	LAS 3	0.833				1.756
	LAS 4	0.843				1.616
	LAS 5	0.866				1.670
	LAS 6	0.829				1.712
	LAS 7	0.839				1.747
	LAS 8	0.772				1.722
	LAS 9	0.765				2.290
Employee Change Readiness	ECR 1	0.732	0.888	0.886	0.740	1.617
	ECR 2	0.661				2.372
	ECR 3	0.782				2.094
	ECR 4	0.893				2.204
	ECR 5	0.762				2.054
	ECR 6	0.815				2.123
	ECR 7	0.708				2.329
	ECR 8	0.566				2.173
	ECR 9	0.685				2.111
Employee Perceived Threat	EPT 1	0.770	0.852	0.853	0.650	1.650
	EPT 2	0.643				1.869
	EPT 3	0.701				1.811
	EPT 4	0.704				1.756
	EPT 5	0.705				1.616
	EPT 6	0.784				1.670
Organizational Support for AI	OSI 1	0.755	0.848	0.849	0.690	1.650
	OSI 2	0.631				1.747
	OSI 3	0.796				1.722
	OSI 4	0.715				2.290
	OSI 5	0.788				1.904
	OSI 6	0.700				2.372
Employee Job Crafting	EJC 1	0.962	0.951	0.951	0.765	2.094
	EJC 2	0.767				2.204
	EJC 3	0.927				2.054
	EJC 4	0.936				2.123
	EJC 5	0.722				2.329
	EJC 6	0.906				2.173

.

The HTMT matrix was computed using Python’s panda’s library, which facilitated efficient calculation and comparison of HTMT correlation ratios. The discriminant validity of the measuring model was evaluated by the Heterotrait–Monotrait Ratio (HTMT) matrix and the Fornell–Larcker criterion, both of which validated the distinctiveness of the constructs. The HTMT matrix, created with the panda’s library, assessed HTMT correlation ratios to confirm that all values were below the 0.85 threshold. The analysis indicated that ECR and LAS had the most excellent HTMT value of 0.758, within acceptable thresholds [69,70], confirming their distinctive characteristics. Furthermore, OSA and ECR demonstrated a valid HTMT score of 0.654. In all paired construct comparisons, HTMT scores satisfied the validity criteria. Each construct’s square root of AVE was compared to inter-construct correlations (off-diagonal elements) in the Fornell–Larcker criterion [69]. Discriminant validity occurs when diagonal values exceed off-diagonal correlations. The analysis confirmed that all constructions met Fornell–Larcker. The square root of ECR’s AVE was 0.682, exceeding its correlations with other constructs like ECR × EJC (0.560) and ECR × LAS (0.769). Employee job crafting (EJC) had a square root of the AVE of 0.875, which was higher than EJC × LAS (0.611) and EJC × OSA (0.441). With diagonal values of 0.702 and 0.795, EPT and LAS showed good discriminant validity, outperforming other constructs. Finally, OSA exhibited a square root of AVE of 0.697, more significant than its associations with ECR and LAS of 0.654 and 0.604, respectively. In

Table 2. Fornell–Larcker criterion and Heterotrait–Monotrait Ratio (HTMT).

Constructs	Employee Change Readiness	Employee Job Crafting	Employee Perceived Threat	Leadership AI Symbolization	Organizational Support for AI
Employee Change Readiness	0.682	0.554	0.648	0.758	0.654
Employee Job Crafting	0.560	0.875	0.438	0.608	0.444
Employee Perceived Threat	0.649	0.436	0.702	0.593	0.601
Leadership AI Symbolization	0.769	0.611	0.597	0.795	0.601
Organizational Support for AI	0.654	0.441	0.601	0.604	0.697

Note: The square root of AVE is presented diagonally and in italics, while the values below correlate with other constructs. Also, HTMT values are depicted above the diagonal values.

, the upper diagonal values indicate HTMT scores, verifying that all ratios are below the threshold of 0.85. The lower diagonal and diagonal values indicate Fornell–Larcker outcomes, wherein the square root of the AVE (diagonal) surpasses the correlations with other components (lower diagonal).

The Panda’s library tabulated variance inflation factor (VIF) data for easy analysis. The VIF values were calculated and tabulated using Python’s panda library, enabling efficient computation and analysis of multicollinearity. To assess the structural equation model’s multicollinearity, the inner model constructions’ VIF was examined. It assesses how much other predictor factors explain a predictor variable in the model [71,72]. VIF levels below 3.3 are usually acceptable, whereas values above 5 suggest multicollinearity. The analysis showed that all inner model connections had VIF values below 3. All VIF values were within acceptable ranges, with ECR → EJC as the highest at 2.372 and OSA × LAS → EPT as the lowest at 1.049. A single-factor Harman test with 35% variance found no common method bias (see

Table 3. VIF inner model.

Relationship	VIF
Employee Change Readiness → Employee Job Crafting	2.372
Employee Perceived Threat →Employee Job Crafting	1.811
Leadership AI Symbolization → Employee Change Readiness	1.811
Leadership AI Symbolization → Employee Job Crafting	2.173
Leadership AI Symbolization → Employee Perceived Threat	1.712
Organizational Support for AI → Employee Perceived Threat	1.722
Organizational Support for AI × Leadership AI Symbolization → Employee Perceived Threat	1.049

). The lack of substantial multicollinearity and common method bias guarantees the reliability and validity of the structural equation model, reinforcing the findings’ robustness.

The SEM fitness was evaluated through various fit indices, such as SRMR, d_ULS, d_G, Chi-square, and NFI. The indices were calculated and organized utilizing the pandas library in Python, as illustrated in

Table 4. Model fitness.

	SRMR	d_ULS	d_G	Chi-Square	NFI
Saturated Model	0.063	3.500	3.000	798.13	0.762
Estimated Model	0.078	4.000	3.800	823.96	0.755

. The pandas library in Python was utilized to organize, compute, and evaluate the model fitness data, facilitating an efficient and precise assessment of SEM. The saturated and estimated models were assessed to confirm their structural adequacy and resilience. The saturated model’s SRMR was 0.063, below 0.08. D_ULS and d_G were 3.500 and 3.000, respectively, indicating acceptable distinction. An acceptable match was indicated by the Chi-square statistic of 798.139 and the NFI value of 0.762. The calculated model had an acceptable fit with an SRMR of 0.078. The Chi-square statistic was 823.96, the NFI was 0.755, and the d_ULS and d_G values increased to 4.000 and 3.800. Both models’ SRMR values < 0.08, and NFI scores confirmed the structural model’s fitness. The robust model fitness indices validated the SEM, affirming the SEM analysis’s validity and the constructs’ proposed relationships.

3.3. Structural Model

The model’s predictive significance was examined utilizing construct R-square and Q-square values. R-square (R²) measures the proportion of variance in the dependent variables explained by the independent variables, while Q-square (Q²) assesses the model’s predictive accuracy using a blindfolding procedure [73].

Table 5. Predictive Relevance.

Constructs	R-Square	Q-Square
Employee Change Readiness	0.591	0.342
Employee Job Crafting	0.472	0.275
Employee Perceived Threat	0.628	0.410

shows that ECR had the highest explanatory power, with an R-square value of 0.591, explaining 59.1% of its variance. Its 0.342 Q-square score confirms its prediction accuracy. EJC had moderate predictive significance, with an R-square value of 0.472 and a Q-square value of 0.275. Last, EPT had a significant R-square value of 0.628 (62.8% variance explained) and a Q-square value of 0.410. The high R² and Q² values validate the structural model’s robustness, affirming the SEM analysis’s validity and the constructs’ proposed relationships.

Hypothesis testing was performed with Python’s statsmodels library, which yielded Beta coefficients, T-statistics, p-values, and confidence intervals for the regression analysis. The data were utilized to assess the importance of the proposed linkages inside the structural equation model. The structural model’s analysis revealed notable direct and indirect relationships, with all hypotheses supported according to Beta, T-statistics, p-values, and confidence intervals (see

Table 6. Path analysis.

Relationship	β	(STDEV)	T-Statistics	p-Values	f-Square	Moderation Effect	CI (Lower, Upper)	Results
H1: Leaders’ AI Symbolization → Job Crafting	0.412	0.073	5.644	0.001	0.190		(0.289, 0.535)	Accepted
H2: Leaders’ AI Symbolization → Change Readiness	0.620	0.061	10.164	0.001	0.384		(0.500, 0.740)	Accepted
H3: Leaders’ AI Symbolization → Perceived Threat (Negative)	−0.372	0.082	4.537	0.001	0.130		(−0.534, −0.210)	Accepted
H4: Leaders’ AI Symbolization → Change Readiness → Job Crafting	0.290	0.056	5.179	0.001	0.150		(0.183, 0.397)	Accepted
H5: Leaders’ AI Symbolization → Perceived Threat → Job Crafting (Negative)	−0.185	0.045	4.111	0.001	0.080		(−0.274, −0.096)	Accepted
H6: Organizational Support × AI Symbolization → Change Readiness	0.550	0.067	8.955	0.001	0.220	Amplified the positive effect (0.55 → 0.65)	(0.436, 0.764)	Accepted
H7: Organizational Support × AI Symbolization → Perceived Threat	−0.380	0.079	3.165	0.002	0.070	Reduced the adverse effect (−0.38 → −0.25)	(−0.489, −0.011)	Accepted
H8: Organizational Support Strengthens Indirect Effect (AI Symbolization → Change Readiness → Job Crafting)	0.290	0.061	5.738	0.001	0.180	Indirect effect increased (0.29 → 0.35)	(0.229, 0.471)	Accepted
H9: Organizational Support Weakens Indirect Effect (AI Symbolization → Perceived Threat → Job Crafting)	−0.185	0.049	2.449	0.014	0.040	Indirect effect decreased (−0.185 → −0.120)	(−0.216, −0.024)	Accepted

). The examination of Hypothesis H1 indicates that leaders’ AI symbolization has a favorable effect on employee job crafting, supported by a substantial Beta (β) value of 0.412, a T-statistic of 5.644, and a p-value of 0.001 (p < 0.05). This association’s confidence interval (CI), ranging from 0.289 to 0.535, excludes zero, thereby validating the hypothesis. The f-square value of 0.190 signifies a moderate effect size, underscoring the practical significance of this relationship within the structural model. These findings support prior studies highlighting the essential impact of leadership behaviors on employee proactivity and flexibility within organizational contexts [74]. The ability of leaders to represent AI integration cultivates a motivated environment, prompting employees to participate in job-crafting activities. In Chinese SMEs, where AI technologies are essential in industries such as healthcare and e-commerce, executives who exemplify AI-driven initiatives serve as role models, encouraging employees’ motivation to redefine and enhance their job responsibilities. In AI-driven healthcare solutions, executives representing AI through the implementation of diagnostic tools can motivate employees to innovate in their responsibilities, enhancing overall organizational adaptability [6].

The examination of Hypothesis H2 validated that leaders’ AI symbolization positively impacts employee change readiness, evidenced by a substantial Beta (β) value of 0.620, a T-statistic of 10.164, and a p-value of 0.001 (p < 0.05). The confidence interval (CI) of 0.500 to 0.740 excludes zero, offering robust statistical evidence for the hypothesis. The f-square value of 0.384 indicates a significant effect size, underscoring the importance of the relationship in the structural model. This result corresponds with the research highlighting the crucial role of leadership in promoting employee readiness for change [36]. This link highlights the significance of symbolic leadership in Chinese SMEs, where technology breakthroughs are swiftly reshaping areas such as e-commerce and healthcare. Leaders in AI-driven e-commerce platforms that demonstrably utilize AI tools for strategic decision making can bolster employees’ trust in embracing new processes and systems, hence promoting employee change readiness [48]. The examination of Hypothesis H3 indicated that leaders’ AI symbolization negatively impacts employees’ perception of threat, evidenced by a significant Beta (β) value of −0.372, a T-statistic of 4.537, and a p-value of 0.001 (p < 0.05). The confidence interval (CI) spans from −0.534 to −0.210 and excludes zero, validating the hypothesis. The f-square value of 0.130 denotes a small to moderate effect size, underscoring the practical significance of this relationship. The finding aligns with studies highlighting that proficient leadership mitigates uncertainty and perceived threats during technology transitions [75,76]. Leaders representing AI adoption can present it as an opportunity rather than a threat, facilitating employees’ comprehension of its advantages and alleviating concerns over displacement or obsolescence. In Chinese SMEs, where AI technologies are revolutionizing sectors such as e-commerce and healthcare, the symbolic acts of leaders significantly impact employees’ impressions. In e-commerce platforms, executives showcasing the benefits of AI in improving consumer personalization and operational efficiency can alleviate employee concerns regarding the threatening aspect of AI [77]. This relationship highlights the essential role of leadership in influencing employees’ perceptions of technology adoption. Symbolic leadership effectively mitigates perceived threats and cultivates an optimal environment for technological innovation, particularly within dynamic and competitive contexts such as Chinese SMEs [51].

The analysis confirmed H4, indicating that leaders’ AI symbolization positively influences employees’ change readiness, subsequently enhancing job crafting. This relationship is characterized by a significant Beta (β) value of 0.290, a T-statistic of 5.179, and a p-value of 0.001 (p < 0.05). The confidence interval (CI) spans from 0.183 to 0.397 and does not encompass zero, reinforcing the hypothesis. The f-square value of 0.150 signifies a moderate effect size, affirming this indirect relationship’s practical significance within the structural model. This finding aligns with theoretical frameworks that associate leadership behaviors with increased employee readiness for change, promoting proactive behaviors such as job crafting [78]. In Chinese SMEs, especially within dynamic sectors such as healthcare and e-commerce, leaders’ symbolic actions enhance employees’ preparedness to adapt to changes introduced by AI-driven workflows and systems. The analysis confirmed H5, indicating that leaders’ AI symbolization can result in perceived threats that negatively affect job crafting, evidenced by a significant Beta (β) value of −0.185, a T-statistic of 4.111, and a p-value of 0.001 (p < 0.05). The confidence interval (CI) spans from −0.274 to −0.096 and does not include zero, thereby affirming the statistical significance of this indirect negative relationship. The f-square value of 0.080 signifies a small yet significant effect size, underscoring the practical implications of this pathway. This finding is consistent with research recognizing possible unintended consequences of leadership behaviors during technological transitions [6]. In e-commerce, employees may perceive AI-driven inventory management systems as competitors rather than tools, adversely impacting job-crafting behaviors [42,76].

The analysis confirmed H6, indicating that organizational support enhances the positive relationship between leaders’ AI symbolization and employee change readiness, evidenced by a significant Beta (β) value of 0.550, a T-statistic of 8.955, and a p-value of 0.001 (p < 0.05). The confidence interval (CI) of 0.436 to 0.764 excludes zero, confirming the moderation effect’s statistical significance. The f-square value of 0.220 signifies a moderate effect size, highlighting the practical significance of this moderating role. The moderation effect enhanced the positive relationship, elevating the influence of AI symbolization on change readiness from 0.55 to 0.65. Organizational support, including providing resources, training, and effective communication, bolsters the credibility of leaders representing AI integration and enhances employees’ preparedness to adjust to technological changes. In healthcare, providing resources for AI-powered diagnostic systems bolsters employees’ confidence in the organization’s commitment, thereby improving their readiness for change [36,54]. The analysis confirmed H7, indicating that organizational support mitigates the negative relationship between leaders’ AI symbolization and employees’ perceived threats. The moderation effect was significant, indicated by a Beta (β) value of −0.380, a T-statistic of 3.165, and a p-value of 0.002 (p < 0.05). The confidence interval (CI) spans from −0.489 to −0.011 and does not include zero, indicating the effect’s statistical significance. The f-square value of 0.070 signifies a small yet significant effect size. The moderation diminished the negative influence of AI symbolization on the perceived threat, altering the effect from −0.38 to −0.25, indicating organizational support’s defensive role. This relationship is significant for supportive organizational practices in alleviating employee concerns and uncertainties during technological transitions. Organizations that offer resources, facilitate clear communication, and provide emotional support enhance employees’ perceptions of AI integration as an opportunity instead of a threat. The analysis confirmed H8, indicating that organizational support enhances the indirect effect of leaders’ AI symbolization on job crafting via employee change readiness. The relationship exhibited a Beta (β) value of 0.290, a T-statistic of 5.738, and a p-value of 0.001 (p < 0.05), indicating statistical significance. The confidence interval (CI) spans from 0.229 to 0.471 and does not include zero, thereby confirming the statistical significance of this moderated indirect effect. The f-square value of 0.180 signifies a moderate effect size, highlighting the significance of this relationship. The moderation enhanced the indirect effect of AI symbolization on job crafting, increasing it from 0.29 to 0.35, thereby illustrating the amplifying influence of organizational support. Organizational support enhances employees’ confidence and equips them to engage in change readiness and job-crafting behaviors, thereby amplifying the influence of leadership. In AI-driven healthcare, offering supplementary training on diagnostic tools can improve employees’ preparedness to adopt AI and adapt their roles to align with new systems. The analysis confirmed H9, indicating that organizational support diminishes the indirect effect of leaders’ AI symbolization on job crafting via employee perceived threat. The relationship demonstrated significance, indicated by a Beta (β) value of −0.185, a T-statistic of 2.449, and a p-value of 0.014 (p < 0.05). The confidence interval (CI) ranges from −0.216 to −0.024 and does not include zero, thereby affirming the significance of the moderated indirect effect. The f-square value of 0.040 signifies a small yet significant effect size. The moderation diminished the negative indirect effect, reducing it from −0.185 to −0.120, illustrating the mitigating influence of organizational support. Supportive practices, including training and resources, mitigate perceived threats and bolster employees’ confidence in adapting to technological innovations.

4. Practical and Theoretical Implications

Our study contributes numerous important theoretical advances to AI in organizational management and job crafting. AI exposure has primarily been shown to improve employee change readiness. However, our data show leaders’ AI symbolization can either benefit or ruin employees. Leaders’ AI symbolization improves change readiness and employee job crafting. According to our latest research, leaders’ AI symbolization negatively impacts employees perceived threat. This dual perspective gives a more balanced and precise view of leaders’ AI symbolization of employees. Our work challenges the unidimensional concept of leaders’ AI symbolization by showing that it can improve employee job crafting. Leaders’ AI symbolization increases change readiness and job crafting. Our study also shows leaders’ AI symbolization can lower employee threat perception. This dual approach helps explain how AI can be integrated into organizations. Our study adds to the literature on leaders’ AI symbolization’s positive and negative effects on organizational management by examining both perspectives. Performance, creativity, and satisfaction at work depend on job crafting. AI requires proactive employee involvement in work and change response. Our work adds to the literature on job crafting by concentrating on leaders’ AI symbolization as an essential component, with the positive influence of change readiness and the negative effect of perceived threat. The study examines mediating and boundary conditions to determine when leaders’ AI symbolization benefits employees. This includes change readiness and the adverse effects of perceived threats, which are novel concepts. Leaders’ AI symbolization can boost change readiness with organizational support for AI as a moderator. This support can boost the indirect impact of leaders’ AI symbolization on job crafting through change readiness. Organizational support can also mitigate the indirect negative impact of leaders’ AI symbolization on job crafting through perceived threats.

Our research has many practical implications for companies adopting AI to improve employee outcomes. Leaders should use AI symbolization to improve staff change readiness and job crafting. Accepting and supporting AI and showing interest in AI technologies can do this. Leaders may promote AI, share their experiences, and urge employees to adopt it. Leadership training should also explain AI’s benefits. Leaders can promote AI and encourage staff positivity with workshops and seminars. Second, leaders should address threats. Transparently describing AI symbolization reduces job insecurity. Leaders should discuss AI implementation’s advantages and cons to raise awareness and address concerns. Communicating and acting honestly about AI can also create employee trust and reduce perceived threats. Third, organizational support can boost change readiness for leaders’ AI symbolization and decrease its perceived threat. Organizations should create a communication plan highlighting AI’s benefits and the necessity for trust and understanding. Staff can participate in updates and forums. Change-related and AI skills training can help employees adjust to new tech. Businesses adopting AI to reduce adverse effects should develop a friendly environment where employees feel appreciated and supported. Feedback, check-ins, and staff engagement programs can do this. Leaders should address and reduce job security and skill relevance concerns among employees. Clear job development and reskilling paths reduce such concerns. Finally, executives should consider the interpretation of AI symbolization by employees. Job crafting can be hampered by employees attributing leaders’ AI symbolization to impression management. Honest and transparent leaders build trust and credibility. Job crafting can be supported through leaders’ AI symbolization by balancing perceived dangers, improving organizational support, and considering employee attributions to integrate AI and create a proactive work environment. Enhancing employee job crafting and reducing negative consequences improves performance, creativity, and job happiness.

5. Conclusions and Future Research Directions

This study explores the dual influence of leaders’ AI symbolization on employees’ job-crafting behaviors in Chinese SMEs. Our findings reveal that while leaders’ AI symbolization enhances employee change readiness and promotes proactive job crafting, it may also unintentionally increase perceived threats, leading to resistance. However, organizational support plays a crucial moderating role by strengthening the positive effects of AI symbolization and mitigating its negative impact. These insights contribute to the growing research on leadership, AI adoption, and employee adaptability in technology-driven environments. This study offers three key takeaways:

• Leadership AI Symbolization Shapes Employee Adaptability—Leaders’ framing of AI significantly influences employees’ perceptions, affecting their willingness to engage in job crafting. This reinforces the importance of symbolic leadership in managing technology-driven transitions.
• Organizational Support as a Strategic Buffer—Organizations that provide employees with resources, training, and psychological safety can enhance positive AI adaptation and mitigate fears associated with automation and job security concerns.
• Context Matters in AI Integration—Chinese SMEs operate under resource constraints, making leader-driven AI messaging even more crucial. Unlike larger firms, where structured support systems exist, SMEs must strategically invest in leadership communication and employee development to maximize the benefits of AI adoption.

This study has several limitations, including its focus on Chinese SMEs and short-term effects. Future research should explore long-term impacts and additional moderating variables, such as employees’ technological literacy. To enhance the generalizability of our findings, we recommend longitudinal studies and cross-cultural comparisons. Investigating the role of employees’ past AI experience and technological literacy could further enrich this field of research.