Influencing Factors and Mechanisms of Value Co-Creation in Artificial Intelligence-Driven Human Resource Management: A System Dynamics Simulation

Abstract

In the era of artificial intelligence, human resource management has undergone significant changes compared to traditional approaches regarding value creation methods and influencing factors. This research aims to utilize grounded theory to comprehensively explore the influencing factors of value co-creation in enterprise human resource management within the context of artificial intelligence. Additionally, this research seeks to capture the dynamic relationships, causal links, and evolutionary patterns among the various elements within the system by constructing a system dynamics model. The results indicated that (1) Environmental factors primarily play a regulatory role, organizational factors serve a supportive role, and participant factors act as the driving force in influencing value co-creation in human resource management. (2) In the context of artificial intelligence, both hardware infrastructure and software capabilities can significantly impact value co-creation in human resource management. This research complements current research on the influencing factors of value co-creation in enterprise human resource management. It offers new perspectives and frameworks for the theoretical development and practical application of value co-creation in this area, supporting companies in effectively managing and developing value co-creation in human resource management.

1. Introduction

Artificial intelligence (AI) is participating in enterprise management in various ways. Especially in the process of enterprise human resource management (HRM), AI has been integrated into various aspects such as HRM strategy [1], human–robot/AI collaboration [2], recruitment, and training [3]. Unlike traditional management models, AI assists enterprises in reducing costs and enhancing efficiency by directly engaging in the value co-creation process of HRM [4]. Research indicates that AI significantly impacts value co-creation through automation, interactivity, and social presence functions [5]. Currently, human resource management value co-creation (HRMVC) under the background of AI has attracted widespread attention in the field of management and is developing towards service science [2,6,7]. Key factors influencing HRMVC include external relationships [8], organizational support [9], and individual engagement [10]. These factors collectively determine how effectively value is created within HRM processes. Compared to HRM practices designed solely by the HRM department, HRMVC can generate greater value [7,11,12]. Therefore, companies are increasingly recognizing the importance of encouraging stakeholders to actively participate in HRMVC, as this is crucial for improving overall organizational performance and employee satisfaction [13].

The concept of value co-creation originally came from the field of marketing, emphasizing the joint creation of value through stakeholder interactions [14,15]. With the successful application of co-creation theory in marketing and service management, scholars have begun to explore its necessity and potential in the field of HRM [5,7,16]. Service-dominant logic (SDL), which views value as co-created through interactions and service exchanges among stakeholders, provides a theoretical foundation for understanding this process in HRM [14,17]. This research posits that the value of HRM practices is defined by stakeholders, and to maximize this value, HRM value should be co-created by both managers and participants [12]. HRMVC is a dynamic and iterative process in which the HRM department collaborates with various stakeholders, such as business managers [18], to collectively address challenges, drive innovation, and implement HRM practices [19]. This collaboration is essential as business managers play a critical role in ensuring that HR strategies align with operational goals, effectively bridging the gap between HR and business needs [20,21].

Before digital technology and AI were widely applied in HRM, traditional HRMVC was primarily influenced by three categories of factors: contextual factors, support factors, and individual factors. Within the SDL framework, contextual factors such as external relationships are crucial as they underpin the stakeholder interactions that drive value co-creation in HRMVC [22,23]. Support factors, such as platform resources, organizational vision and values [12], team trust atmosphere [13], organizational change [12], and the technological infrastructure, are essential under the SDL framework as they facilitate the integration of resources necessary for value co-creation in HRMVC. Individual factors, such as individual needs [24], characteristics [25], participation motivation [22], psychological safety, and interactions between participants [7], are pivotal as they influence the quality of stakeholder interactions that drive value co-creation in HRMVC.

According to the SDL framework, the introduction of digital technology and the application of AI represent a significant expansion of the resources available for value co-creation in HRMVC [26,27]. First, AI has become a new entity in value co-creation. It not only participates in daily HRM practices but also supports decision-making through data analysis and prediction, therefore expanding the dimensions and depth of value co-creation [2,28]. Moreover, the integration of AI into HRM systems facilitates a more holistic approach to employee engagement by leveraging system dynamics and feedback loops, which enhance organizational adaptability and resilience [1]. Additionally, the AI readiness of organizations and individuals has become a new critical influencing factor, including the development of technical infrastructure [29], the cultivation of AI-related talent [3,30], and cultural adaptation [31]. AI technology enhances the efficiency of HRM practices through automation and intelligent services while also driving the restructuring of organizational management systems, such as optimizing recruitment and performance evaluation processes through intelligent algorithms [28,32]. Furthermore, AI strengthens the ecological network value of HRM, enabling organizations to better connect and collaborate with external resources [13].

Despite the growing use of digitalization and AI technology in HRMVC most current research primarily focuses on the technical application and impact of these digital technologies and AI within enterprise HRM [2,24,33]. Systematic research on how they contribute to value co-creation in different types of organizations is relatively scarce. However, the role of AI in shaping value co-creation mechanisms within HRMVC, particularly through its influence on stakeholder interactions and resource exchanges, remains underexplored. Regarding the factors influencing HRMVC, scholars have gradually paid attention to contextual and organizational factors such as labor-management and government relations [30], corporate values [34], corporate culture [35], and technology [33], as well as individual-level factors such as psychological safety [36], competence trust [37], and knowledge skills [21]. However, there is a lack of understanding of how AI alters these factors’ roles and their dynamic interplay within HRMVC. Given the importance of stakeholder interactions and resource integration in the SDL framework, there is an urgent need for more systematic research on the role of AI in HRMVC, particularly in understanding how AI-driven resource exchanges and interactions reshape traditional value co-creation processes [38]. Furthermore, existing research mostly relies on quantitative methods such as questionnaire surveys to explore the impact of one or several factors on HRMVC, further limiting insights into the complex, interconnected nature of these processes.

In this context, the research objectives are to systematically explore the dimensions of HRMVC, identify and analyze the influencing factors, and understand the interactions among these factors within the HRM framework. This research is grounded in the theory of value co-creation, using the grounded theory method to systematically identify and categorize the influencing factors of HRMVC, followed by the application of system dynamics to simulate and analyze the interactions among these factors. To achieve these objectives, this research addresses the following research questions:

(1)

What dimensions does HRMVC encompass?

(2)

What are the factors influencing HRMVC? What are the relationships between the influencing factors and HRMVC, as well as among the influencing factors themselves?

(3)

In the context of AI, how do different combinations of AI-related influencing factors affect HRMVC?

This research enhances the understanding of HRMVC by investigating the integration of AI into value co-creation processes, thus contributing to the theoretical advancement of HRM. Additionally, the findings offer practical insights that support companies in implementing digital HRM strategies and improving value co-creation, therefore improving organizational effectiveness and employee engagement.

2. Identification of Influencing Factors

2.1. Research Design

Grounded theory is a research method that builds theory based on empirical data and a deep understanding of phenomena. This method emphasizes generating theories through the systematic collection and analysis of data rather than validating preconceived concepts or assumptions. The main advantage of grounded theory lies in its in-depth exploration of phenomena, which can reveal underlying influencing factors and mechanisms [39]. Grounded theory involves three primary stages: data collection, data coding, and theory development. Initially, researchers gather data through empirical observations and interviews related to the phenomenon under investigation. Subsequently, they conduct open coding, axial coding, and selective coding on the collected data to identify patterns and relationships [40]. The grounded theory is widely used in the field of management studies to explore “what” and “how”, making it suitable for studying the influencing factors of HRMVC and the relationships between these factors.

2.2. Data Sources

This research obtained research data through two channels: interview survey and non-interview survey. The interview survey data, serving as primary sources, were obtained from direct interviews and questionnaire surveys conducted with multiple companies involved in HRMVC activities between 2021 and 2023. The aim was to directly gather insights into the factors influencing HRMVC from senior executives, HRM professionals, employees, and other HRMVC-related personnel. Non-interview survey data, serving as secondary sources, mainly come from company websites, internal documents provided by companies, government documents, case materials, online reports, and literature sources. These data provide important supplements and verification for the interview survey data. After removing irrelevant information and standardizing terminology, we obtained a total of 446 samples (see

Table 1. Summary of research data.

Data Type		Data Sources	Sample Size	Coding
Interview survey data	Enterprise Interview	Online and offline interviews, on-site investigations, WeChat consultations, re-interviews with relevant participants about the company introduction, etc.	33	I1
	Questionnaire survey	Questionnaire data.	217	I2
Non-interview survey data	Government documents	Laws and regulations, policy compilations, government portals, etc.	34	N1
	Enterprise website	Chint Group, Huawei Group, and other enterprises’ official websites.	18	N2
	Internal enterprise documents	Internal enterprise documents provided by respondents.	12	N3
	Case materials	Obtain HRMVC-related case materials from the China Management Case-sharing Centre and The Global Platform of China Cases.	3	N4
	Online reports	Baidu search and WeChat search.	67	N5
	Academic Literature	Chinese National Knowledge Infrastructure, Web of Science, Google Scholar, and Scopus.	62	N6
Total			446	-

). Among these, 296 samples were randomly selected as preliminary coding samples, with 150 samples reserved for subsequent model saturation testing.

2.2.1. Interview Survey Data

The interview survey data consists of the following two parts:

(1)

Enterprise interview materials. During the period from 2021 to 2023, we conducted semi-structured interviews through online and offline visits, as well as WeChat consultations, with various enterprises, including Chint Electric, NewMed Medical, DSM China, Otis Elevator, KingMed Diagnostics, and Freshippo robot restaurants. The interviews mainly focused on the interaction process between human resource managers and stakeholders during the HRMVC process. To comprehensively understand the influencing factors of HRMVC, we selected senior executives, human resources professionals, business department leaders, and company employees as interview subjects, following the principles of theoretical sampling and data availability. To ensure that the selected interviewees are sufficiently representative and diverse while also considering the impact of factors such as differences in age structure and job structure, we established the following selection criteria for interview subjects: ① The interview subjects are related to HRMVC work. ② The gender distribution of the interview subjects is balanced. ③ The job structure distribution of the interview subjects is reasonable. Ultimately, a total of 24 interview subjects were selected. Among them, 50% were male. Regarding job structure, 12.5% were senior executives, 16.6% were business department leaders, 25% were human resources staff, and 45% were company employees. During the interview, we first explained the concept of HRMVC to ensure their understanding. Then, we conducted in-depth interviews based on an open-ended questionnaire. The main interview questions included: What HRMVC situations exist in the company? How are these activities carried out? What factors can influence HRMVC? In which part of the HRMVC work have you mainly participated?

(2)

Enterprise questionnaire sample. We collected questionnaire data using an online platform, obtaining a total of 245 samples, of which 217 were valid. The questionnaire items included questions related to HRMVC, such as “What HRMVC situations exist in the company?” and “What factors can influence HRMVC?”

2.2.2. Non-Interview Survey Data

The non-interview survey data for this research primarily involved the following six aspects:

(1)

Government documents. We selected government documents related to HRMVC from human resources laws and regulations. After screening and analysis, a total of 34 government documents were included in the sample.

(2)

Enterprise website. Based on the corporate websites with HRMVC-related information, 18 documents were selected and included in the research sample.

(3)

Internal enterprise documents. Interviewees provided 12 documents related to HRMVC, including notifications, plans, and other relevant materials.

(4)

Case materials. We obtained 3 teaching case studies related to HRMVC themes from the China Management Case-sharing Centre and The Global Platform of China Cases.

(5)

Online reports. Using “HRMVC” and its similar expressions as search terms, we conducted a search through online platforms, carefully selecting authoritative reports closely related to the topic and excluding low-relevance samples. Ultimately, 67 online report samples were included in the coding analysis.

(6)

Academic Literature. A total of 62 relevant academic articles were collected from Chinese National Knowledge Infrastructure, Web of Science, Google Scholar, and Scopus.

2.3. Data Coding

2.3.1. Open Coding

Open coding is a process that begins with raw data, employing comparative methods to identify similarities and differences. This allows for the assignment of concepts that describe phenomena, which are then categorized to gradually synthesize the initial data. To facilitate effective open coding, researchers engage in multiple readings of the raw materials, becoming thoroughly familiar with and deeply understanding the content before altering the original order of the materials. The coding process does not involve predefined concepts; rather, it is entirely data-driven. During the open coding, we conducted a detailed analysis of the raw data, identifying key concepts and phenomena. Ultimately, we identified 474 original statements and corresponding initial concepts. Due to the presence of some repetition and overlap in the initial concepts, we only selected initial concepts that appeared more than three times. Ultimately, 51 concepts and 14 categories (marketing environment, policy environment, employment environment, organizational environment, HRM, co-creation support, individual characteristics, social identity, future expectations, knowledge and skills, strategic value co-creation, economic value co-creation, capability value co-creation, and platform value co-creation) were refined. The results of open coding are shown in

Table 2. The results of open coding.

Original Material	Open Coding
	Phenomenon	Conceptualization	Categorization
After COVID-19, market demand has not yet recovered, so enterprises are facing more severe competition. At such times, it requires us to pay more attention to the overall costs of enterprises, and achieve cost reduction and efficiency improvement. Regarding HRM, how to use machines and human resources to optimize productivity more effectively requires leadership from company executives. HRM departments and production departments need to collaborate to calculate and plan this matter together. (I1)	a1: market demand did not recover after the COVID-19 a2: productivity allocation a3: senior management takes the lead, HRM department and production department cooperate to complete the work	aa1: economic environment (a1) aa2: demand coordination (a2) aa3: relationship management (a3)	A1: marketing environment (aa1) A2: co-creation support (aa2, aa3)
I think the company’s atmosphere is quite good, and its future development also looks promising. Therefore, when the company has job openings, and if I happen to have friends looking for jobs, I am more willing to recommend my friends to come and work here. (I1)	a14: the company atmosphere is good a15: the company is developing well	aa14: organizational climate (a14) aa15: organizational identification (a15)	A6: organizational environment (aa14) A7: social identity (aa15)
The company policy states that if an employee’s suggestion is adopted, there will be an additional bonus. Well, how should I put it? Personally, I also have an outgoing personality and enjoy getting involved in these things. (I1)	a7: the company has a reward system a17: extrovert personality	aa5: system design (a7) aa9: personality traits (a17)	A4: individual characteristics (a17) A5: HRM (aa5)

Note: Due to the large amount of encoded content, this table only represents a selected portion of categories for display purposes.

.

2.3.2. Axial Coding

Axial coding involves establishing connections between the categories derived from open coding through clustering analysis. It groups categories with similar themes based on their logical relationships and interconnections to form main categories. This research adopts the paradigm model from grounded theory, identifying and establishing relationships among categories from four aspects: causal conditions, contextual/intervening conditions, action/interaction strategies, and consequences [41]. As a result, 4 main categories, 13 initial categories, and 40 corresponding categories were e summarized. The content of each main category and its corresponding categories are shown in

Table 3. The result of axial coding.

Main Category	Initial Category	Corresponding Category	Category Interpretation
Environmental factor	Marketing environment	Economic environment	The market economy situation of the enterprise.
		Product competitiveness	The market share and sales performance of the enterprise products.
	Policy environment	Laws and regulations	Relevant labor laws, local regulations, and other mandatory requirements.
		Policy document	Government subsidies, social security, taxation, and other relevant operation or guidance documents.
		Work efficiency	The efficiency of the interface window or office with the government.
	Employment environment	Employment situation	The employment status of the area in which the enterprise is located.
		“Human + AI” mixed labor force	Both traditional labor and AI are put into production as labor that can create value.
Organizational factor	Organizational environment	Enterprise characteristic	The industry and scale of the enterprise.
		Enterprise reputation	Enterprise industry reputation, network evaluation, legal disputes, etc.
		Organizational climate	The overall atmosphere that participants feel in the organization.
		Organizational justice	A participant’s perception of fairness in organizational distribution, procedures, etc.
		Digitization level	The extent to which enterprises use big data and AI to improve production and management.
		Innovation competence	Enterprise’s ability to create and improve products and management.
	HRM	System design	Arrangement of HRM process and system.
		Policy training	Enterprises train relevant personnel on human resources policies and procedures.
		Executive feedback	Constantly collect participants’ opinions for improvement during HRM system implementation.
	Co-creation support	Resource platform	Enterprises use hardware, personnel, and software to build the HRMVC platform.
		Demand coordination	Communicate and coordinate the needs of participants.
		Leadership support	HRMVC leadership support for co-creation efforts.
		AI empowerment	The organization provides AI-related hardware and software support to help participants work.
		Relationship management	Organize and coordinate the relationships and interactions of various participants.
Participant factor	Individual characteristics	Demographic characteristics	Gender, age, and other characteristics of the participants.
		Personality traits	Participants had different personalities, such as introversion and extroversion.
		Psychological security	Participants feel safe in the organization to advise and try new things.
	Social identity	Career identity	The degree to which the participants recognized their occupation.
		Organizational identification	The degree to which participants recognize the organization.
	Future expectations	Economic expectation	Participants’ judgments about the future economic situation.
		Enterprise expectation	Participants’ judgment on the future development situation of the enterprise.
		Personal planning	Participants’ plans for their own futures.
	Knowledge and skills	Professional knowledge and skills	The degree to which the participants have mastered and applied their professional knowledge to the job.
		Basic knowledge of human resources	Participants’ knowledge of basic knowledge of human resources.
		Management communication ability	Management ability, collaboration and communication ability of participants.
		Innovation ability	The ability of participants to identify problems, create and improve.
		AI capability	The ability of participants to work with AI technology.
HRMVC	Value co-creation environment	Organizational external environment	Environmental factors outside the organization that affect HRMVC.
		Internal organizational environment	HRM policy, leadership support, and other internal environment within the organization.
	Value co-creation willingness	Organizational co-creation willingness	The extent to which the organization is willing to undertake HRMVC.
		Individual co-creation intention	The degree to which individuals are willing to participate in HRMVC.
	Value co-creation capacity	Organizational co-creation ability	The ability of the organization to support HRMVC activities.
		Individual co-creation ability	The ability of the individual to participate in HRMVC and achieve the corresponding goals.

.

2.3.3. Selective Coding

Selective coding builds upon the findings from axial coding, organizing the relationships among key categories into a coherent narrative to further develop a new theoretical framework. By reviewing the literature and analyzing interview data, and through iterative comparison and induction, this research identifies the core category as “Factors Influencing Enterprise HRMVC”. The internal logic and relationships between each category are detailed in

Table 4. Typical relational structures among the main categories.

Typical Relation Structure	Connotation of Relation Structure
Environmental factor → HRMVC	The marketing environment, policy environment, and employment environment in which an enterprise operates will impact the company’s HRMVC.
Organizational factor → HRMVC	The internal environment of the enterprise and HRM, as well as the support it provides to HRMVC, will have a direct impact.
Participant factor → HRMVC	Individual characteristics, social identity, future expectations, and knowledge and skills of HRMVC participants influence HRMVC.
Environmental factor → Organizational factor → HRMVC	The external environment, including markets, policies, and talent, can influence a company’s environment and HRM policies. These factors also affect the co-creation support that a company can provide, therefore impacting HRMVC.
Environmental factor → Participant factor → HRMVC	The external economic and employment environment influences HRMVC through factors such as the social identity and future expectations of its participants, therefore impacting HRMVC.
Organizational factor → Participant factor → HRMVC	Organizations influence HRMVC by supporting and managing participants’ organizational identification, future expectations, and knowledge skills through HRM policies and provided resources.

. The analysis reveals that the 10 initial categories all have a significant impact on enterprise HRMVC. On this basis, the influencing factors of the 10 initial categories are further summarized into 3 main categories: environmental factor, organizational factor, and participant factor (see Figure 1 for details).

2.3.4. Theoretical Saturation Test

When no new concepts, categories, or relationships emerge in the reserved samples, it is considered that the concepts or categories identified in the research have covered all the data that can be obtained from the samples. This can be regarded as theoretical saturation. We found that the theoretical model was essentially saturated after the initial coding of 296 samples. An additional 150 reserved samples were then coded, revealing no new concepts, categories, or relationships between categories. This indicates that all categories in the HRMVC influencing factors model were fully explored. Therefore, the enterprise HRMVC influencing factors model framework obtained in this research has passed the theoretical saturation test.

3. Simulation Analysis

A system dynamics model can describe a system’s behavior over time, capturing dynamic and causal relationships, as well as the evolutionary patterns among its internal elements. It is widely utilized to study the interactions and feedback mechanisms among various elements within a system. The enterprise HRMVC is a complex, open, dynamic system that involves environmental factors, participant factors, and organizational factors, all of which interact and collectively influence the enterprise HRMVC. This system exhibits characteristics such as nonlinearity, multivariable interactions, complexity, dynamism, and openness. Therefore, this research employs the Vensim PLE 10.2 to construct the enterprise HRMVC system, demonstrating its internal structure and complex behavior through simulation and emulation, which helps to understand the essence and intrinsic connections of the enterprise HRMVC system.

3.1. Determination of System Boundaries

The enterprise HRMVC system comprises the organization subsystem, the environment subsystem, and the participants subsystem. These three subsystems interact with and influence each other, jointly affecting the three dimensions of the HRMVC process: value co-creation environment, value co-creation capability, and value co-creation willingness. The behavioral subjects in the enterprise HRMVC process include the government, the enterprise HRM departments, senior management, business managers, employees, and other multiple entities. These actors enhance the value of each subsystem through internal value flow and external value creation, forming a dynamic development process of value co-creation. The main hypotheses of the enterprise HRMVC system are as follows:

(1)

The evolution of enterprise HRMVC is a continuous and gradual development process.

(2)

Other unexpected events or force majeure factors that may impact the system evolution, such as public health emergencies, wars, and natural disasters, are excluded.

(3)

The enterprise HRMVC system is a relatively stable system, primarily influenced by organizational, environmental, and participant factors.

3.2. Analysis of Systematic Causality

Causal loop diagrams are the foundation of system dynamics modeling. They illustrate and analyze the interactions between internal variables in a system by depicting the causal relationships among these variables. Causal loop diagrams typically consist of nodes and directed arrows: nodes represent the various variables within the system, and directed arrows indicate the causal relationships between variables. The direction of the arrow shows the influence or effect of the starting variable on the ending variable. Directed arrows include positive feedback loops (+) and negative feedback loops (−). The causal loop diagram of the enterprise HRMVC system is detailed in Figure 2, and this system mainly includes three positive feedback loops:

(1)

Value Co-creation Environment $\stackrel{+}{\to }$ HRMVC $\stackrel{+}{\to }$ Co-creation Support $\stackrel{+}{\to }$ Value Co-creation Environment

(2)

Value Co-creation Willingness $\stackrel{+}{\to }$ HRMVC $\stackrel{+}{\to }$ Co-creation Support $\stackrel{+}{\to }$ Value Co-creation Willingness

(3)

Value Co-creation Capacity $\stackrel{+}{\to }$ HRMVC $\stackrel{+}{\to }$ Co-creation Support $\stackrel{+}{\to }$ Value Co-creation Capacity

3.3. Analysis of System Flow Diagram

System stock-flow diagrams are based on causal loop diagrams, which depict the dynamic behavioral patterns of the stock variables and flow variables within the system over time. This helps to reveal the dynamic characteristics, delay effects, and feedback mechanisms of the system, providing a deeper understanding of the internal structure and dynamic evolution. The stock and flow diagram of the enterprise HRMVC system is shown in Figure 3.

This research involves 3 state variables, 3 rate variables, 33 auxiliary variables, and 28 constants, as detailed in Appendix A

Table A1. Description of the HRMVC system variables.

Variable Classification	Variable Name	Variable Description
State variable	Value co-creation environment	The external environmental factors and internal environmental factors that influence enterprise HRMVC
	Value co-creation capacity	The combination of various abilities that participants need for the HRMVC event
	Value co-creation willingness	Participants’ willingness to participate in HRMVC
Rate variable	The increase in value co-creation environment	To measure the added value of a value co-creation environment
	The increase in value co-creation capacity	To measure the added value of a value co-creation capacity
	The increase in value co-creation willingness	To measure the added value of a value co-creation willingness
Auxiliary variable	HRMVC	HRM stakeholders jointly participate in creating the value of HRM activities
	Policy environment	Government-related factors such as laws and regulations, normal documents, and work efficiency
	Marketing environment	Economic environment, product competitiveness, and other market economic factors
	Employment environment	Employment situation and labor force structure and other enterprise employment factors
	Employment situation	General labor market conditions and competition for job seekers
	Economic environment	The market economy situation of the enterprise
	Organizational environment	Characteristics, atmosphere, reputation, and other conditions of the enterprise
	Innovation competence	Enterprise’s ability to create and improve products and management
	Co-creation support	Resource platform, requirement coordination, and other support provided by the organization for HRMVC
	HRM	HRM system design, policy training, and other management work carried out by enterprises
	Knowledge and skills	The knowledge and skill base necessary for participants to participate in HRMVC activities
	Laws and regulations	Relevant labor laws, local regulations, and other mandatory requirements
	Individual characteristics	Participants’ personality, psychological safety, and other factors
	Future expectations	Participants’ expectations for the future of the economy, companies, and individuals
	Social identity	The participants’ level of recognition of their profession and organization
	Demographic characteristics	Gender, age, and other characteristics of the participants
	GDP	The final results of production activities of all permanent resident units in a country over a certain period of time
	Per capita GDP	Per capita gross domestic product
	The difference between imports and exports of goods	Gross export—gross import
	The debt balance of the central government	Government debt—debts paid
	Consumer price index	The per capita consumption expenditure of permanent residents on purchasing and using goods and services in domestic and international markets to directly meet their living needs
	Per capita disposal income	Consumer price index/resident population
	Foreign direct investment	Foreign investors in our country invest through establishing foreign-invested enterprises, partnerships, joint exploration and development of petroleum resources with Chinese investors, and setting up branches of foreign companies
	Number of labor dispute arbitration cases accepted	The Arbitration Committee for Labor Disputes and Personnel Disputes, in accordance with national laws, regulations, and relevant rules and regulations, reviews the arbitration applications submitted by the parties involved in labor disputes and personnel disputes. The committee formally registers the number of labor disputes and personnel dispute cases that meet the acceptance criteria after examination
	Labor union	The number of labor unions
	Income from social insurance funds	The funds formed from insurance premiums paid by units participating in social old-age, unemployment, medical, maternity, and work injury insurance, according to the national regulations on payment base and contribution rates, as well as funds acquired through other legal means
	Labor force population	The population aged 16 and above, capable of working, participating in, or requesting participation in socio-economic activities, including employed and unemployed individuals
	Employed person	A laborer who is at least 16 years old and works for compensation or profit
	Registered urban unemployment rate	The percentage of the urban unemployed population to the sum of the urban employed population and unemployed population
	Average wage	The average wage earned per person employed in a certain period
	The number of authorized patents	The number of authorized patents held by the company both domestically and internationally
	Number of patent applications accepted	The number of domestic and international patent applications accepted
	Population age structure	The proportion of the population aged 15–64 years old to the total population
Constant	Product competitiveness	The market share and sales performance of the enterprise products
	Policy document	Government subsidies, social security, taxation, and other relevant operation or guidance documents
	Work efficiency	The efficiency of the interface window or office with the government
	“Human + AI” mixed labor force	Both traditional labor and AI are put into production as labor that can create value
	Enterprise characteristic	The industry and scale of the enterprise
	Enterprise reputation	Enterprise industry reputation, network evaluation, legal disputes, etc.
	Organizational climate	The overall atmosphere that participants feel in the organization
	Organizational justice	A participant’s perception of fairness in organizational distribution, procedures, etc.
	Digitization level	The extent to which enterprises use big data and AI to improve production and management
	System design	Arrangement of HRM process and system
	Policy training	Enterprises train relevant personnel on human resources policies and procedures
	Executive feedback	Constantly collect participants’ opinions for improvement during HRM system implementation
	Resource platform	Enterprises use hardware, personnel, and software to build the HRMVC platform
	Demand coordination	Communicate and coordinate the needs of participants
	Leadership support	HRMVC leadership support for co-creation efforts
	AI empowerment	The organization provides AI-related hardware and software support to help participants work
	Relationship management	Organize and coordinate the relationships and interactions of various participants
	Personality traits	Participants had different personalities, such as introversion and extroversion
	Psychological security	Participants feel safe in the organization to advise and try new things
	Career identity	The degree to which the participants recognized their occupation
	Organizational identification	The degree to which participants recognize the organization
	Economic expectation	Participants’ judgments about the future economic situation
	Enterprise expectation	Participants’ judgment on the future development situation of the enterprise
	Personal planning	Participants’ plans for their own futures
	Professional knowledge and skills	The degree to which the participants have mastered and applied their professional knowledge to the job
	Basic knowledge of human resources	Participants’ knowledge of basic knowledge of human resources
	Management communication ability	Management ability, collaboration and communication ability of participants
	Innovation ability	The ability of participants to identify problems, create and improve

. Among these, data for 17 constants were sourced from the “China Statistical Yearbook” and processed using 0–1 standardization. The simulation spanned 15 years with a timestep of 1 year, starting from INITIAL TIME = 2008 to FINAL TIME = 2022, using TIMESTEP = 1. To establish functional equations, this research utilized the expert scoring method to determine the weights of each variable. Scores were collected from various groups, including university professors, government officials, corporate executives, and mid-level managers, yielding a total of 20 score sets. The final weights were averaged to establish the variable weights, with detailed equations provided in Appendix B.

3.4. Simulation Analysis of Influence Mechanism

To provide an intuitive depiction of the internal structure and dynamic visualization of the enterprise HRMVC system and to gain a comprehensive understanding of the complexity and patterns of its changes, this research examines the development trends of the enterprise HRMVC system through the design of three different simulation scenarios: single-factor influence mechanism, multi-factor synergistic influence mechanism, and multi-factor non-synergistic influence mechanism.

3.4.1. Single-Factor Influence Mechanism

This research primarily simulates the influence mechanisms of organizational factors, environmental factors, and participant factors on the enterprise HRMVC system by controlling changes in the values of normal variables. The simulation scheme for the single-factor influence mechanism is shown in

Table 5. Single-factor influence mechanism simulation schemes.

Simulation Schemes	Organizational Factors	Environmental Factors	Participant Factors
Scenario 1	↑10%
Scenario 2	↑20%
Scenario 3		↑10%
Scenario 4		↑20%
Scenario 5			↑10%
Scenario 6			↑20%

, and the simulated evolution trend is illustrated in Figure 4.

Comparing the HRMVC curves of the six simulation scenarios in Figure 4, it is evident that all six scenarios have a significant positive impact on enhancing enterprise HRMVC. In terms of the effectiveness of the simulation scenarios, Scenario 4 > Scenario 2 > Scenario 6 > Scenario 1 > Scenario 3 > Scenario 5. From this, it can be concluded that the strength of government policy regulation, including a favorable policy environment and employment environment, has the most significant impact on enterprise HRMVC. Following this is the level of organizational support for HRMVC, and then participants’ future expectations and job skills. Therefore, when resources are limited, priority should be given to enhancing government policy regulation, supplemented by increased organizational support for HRMVC and improving participants’ future expectations and job skills. This approach can maximize the enhancement of enterprise HRMVC behavior and levels.

3.4.2. Multi-Factor Synergistic Influence Mechanism

This research focuses on the influencing factors and mechanisms of HRMVC in the context of AI. Given the numerous influencing factors within the system, the investigation of multi-factor synergistic influencing mechanisms primarily focuses on variables related to AI, including “human + AI” mixed labor force, digitization level, AI empowerment, and AI capability. The simulation schemes for the synergistic influencing mechanisms involving multiple factors are presented in

Table 6. Multi-factor synergistic influence mechanism simulation schemes.

Simulation Schemes	“Human + AI” Mixed Labor Force	Digitization Level	AI Empowerment	AI Capability
Scenario 1	↑10%	↑5%	↑5%	↑5%
Scenario 2	↑20%	↑5%	↑5%	↑5%
Scenario 3	↑5%	↑10%	↑5%	↑5%
Scenario 4	↑5%	↑20%	↑5%	↑5%
Scenario 5	↑5%	↑5%	↑10%	↑5%
Scenario 6	↑5%	↑5%	↑20%	↑5%
Scenario 7	↑5%	↑5%	↑5%	↑10%
Scenario 8	↑5%	↑5%	↑5%	↑20%

, and the evolutionary trend in simulation is depicted in Figure 5.

By comparing the curves in Figure 5, it was found that different combination schemes have a significant positive impact on the overall level of HRMVC. However, the effectiveness of each scheme varies, with Scenario 6 > Scenario 2 > Scenario 5 > Scenario 4 > Scenario 1 > Scenario 8 > Scenario 3 > Scenario 7. From this, it can be concluded that when resources are limited, priority should be given to empowering participants in enterprises using AI while increasing the use of “human + AI” mixed labor and introducing AI as a new type of labor to participate in value creation, therefore enhancing the digital intelligence level of enterprises and consequently improving HRMVC levels. Furthermore, suppose participants fully understand the trend and efficiency of AI involvement in work, and enhance AI capabilities through learning, therefore improving the efficiency of human-AI collaboration as a new type of labor force. In that case, it can also enhance enterprise HRMVC levels.

3.4.3. Multi-Factor Non-Synergistic Influence Mechanism

In enterprise practice, HRMVC systems exhibit various complex situations where multiple variables interact with each other, albeit not entirely in a coordinated manner. This implies that the relationships among variables are neither simply linear nor completely independent. Therefore, this research sets up multiple simulation scenarios to investigate the multi-factor non-synergistic influence mechanism of enterprise HRMVC systems, as detailed in

Table 7. Multi-factor non-synergistic influence mechanism simulation schemes.

Simulation Schemes	“Human + AI” Mixed Labor Force	Digitization Level	AI Empowerment	AI Capability
Scenario 1	↑20%		↑5%	↑5%
Scenario 2	↑20%	↑5%		↑5%
Scenario 3	↑20%	↑5%	↑5%
Scenario 4		↑20%	↑5%	↑5%
Scenario 5	↑5%	↑20%		↑5%
Scenario 6	↑5%	↑20%	↑5%
Scenario 7		↑5%	↑20%	↑5%
Scenario 8	↑5%		↑20%	↑5%
Scenario 9	↑5%	↑5%	↑20%
Scenario 10		↑5%	↑5%	↑20%
Scenario 11	↑5%		↑5%	↑20%
Scenario 12	↑5%	↑5%		↑20%

and Figure 6.

The analysis shows that all the schemes can have a positive impact on HRMVC, specifically:

(1)

Scenario 3 > Scenario 1 > Scenario 2. In enterprises adopting a “human + AI” hybrid workforce while enhancing the level of digital intelligence, using AI to empower participants and improve their AI capabilities can enhance the enterprise’s HRMVC level. Enterprises focusing on enhancing the level of digital intelligence and using AI to empower participants in their work yield better results compared to scenarios where enterprises empower participants while focusing solely on enhancing their own AI capabilities. Both of these scenarios outperform enterprises that only focus on overall digital intelligence without empowering AI for specific work tasks.

(2)

Scenario 6 > Scenario 4 > Scenario 5. It can be seen that when enterprises emphasize the improvement of digital intelligence level, the combined effect of “human + AI” mixed workforce and empowering participants with AI is better than the effect of enterprises empowering AI while participants focus on improving their own AI capabilities. These two schemes are more effective than enterprises using a “human + AI” mixed workforce while participants focus on enhancing their own AI capabilities.

(3)

Scenario 8 > Scenario 9 = Scenario 7. It can be seen that after enterprises empower participants with AI and use a “human + AI” hybrid workforce, schemes focusing on enhancing participants’ own AI capabilities are superior to situations where enterprises enhance their digitalization level and adopt a “human + AI” hybrid workforce, as well as cases where enterprises enhance their digitalization level while participants improve their own AI capabilities. Therefore, it can be seen that after enterprises empower participants with AI, the use of a “human + AI” hybrid workforce is more effective than schemes focusing solely on enhancing enterprise digitalization or improving participants’ AI capabilities.

(4)

Scenario 11 > Scenario 10 > Scenario 12. It can be seen that when participants focus on improving their own AI capabilities, the effect of enterprises using a “human + AI” mixed workforce while empowering participants with AI is better than enterprises focusing on improving their own digital intelligence level while empowering participants with AI. Both are more effective than enterprises using a “human + AI” mixed workforce while simultaneously focusing on improving their own digital intelligence level.

Overall, the analysis of the simulation results for the multi-factor non-synergistic influence mechanism shows that Scenario 8 > Scenario 9 = Scenario 7 > Scenario 3 > Scenario 1 > Scenario 2 > Scenario 6 > Scenario 4 > Scenario 11 > Scenario 5 > Scenario 10 > Scenario 12 > Initial value. This demonstrates that in various scenarios of resource allocation and interactions, the overall effect of empowering participants with AI is superior to using a “human + AI” mixed.

4. Discussion

This research employs grounded theory and system dynamics methods to systematically explore the influencing factors and mechanisms of HRMVC and has primarily reached the following conclusions:

(1)

Based on the varying content of co-creation, this research categorizes HRMVC into three dimensions: value co-creation environment, value co-creation capability, and value co-creation willingness. These three dimensions collectively constitute important factors influencing the effectiveness of enterprise HRMVC activities. ① The dimension of value co-creation environment encompasses both the external and internal environments in which a company operates. ② Value co-creation capability represents the combination of various skills and resources required by participants in HRMVC activities, including expertise, communication skills, coordination abilities, and resource integration capabilities. ③ Value co-creation willingness refers to participants’ enthusiasm and motivation towards HRMVC activities. The strength of participants’ willingness directly impacts their level of engagement in the co-creation process.

(2)

This research constructs a model of influencing factors for HRMVC through grounded theory analysis, identifying 10 major influencing factors categorized into three main groups: environmental factors, organizational factors, and participant factors. These findings align with previous research that emphasizes the role of environmental and organizational factors in HRM [22]. In addition to factors previously identified in the literature, our research reveals additional influencing factors, such as AI empowerment at the organizational level and AI capability among participants. We have further integrated all these factors into a comprehensive model, providing a more systematic and complete representation of HRMVC factors, therefore enhancing the understanding of their interactions within the HRMVC framework.

(3)

Based on the HRMVC influencing factors model, this research uses system dynamics simulation analysis and finds that, under single-factor conditions, environmental factors such as government policies and employment situations have a significant positive impact on enterprise HRMVC. Additionally, comprehensive support from the company for HRMVC activities also plays an important role. This confirms existing studies that highlight the importance of external and organizational factors [12]. However, our findings extend the existing literature by demonstrating how these factors specifically influence HRMVC within the context of AI, which offers new insights into the dynamic interactions of these factors under AI-driven conditions, a perspective that previous studies have not fully explored.

(4)

In the context of AI, this research focuses on multi-factor non-synergistic and synergistic simulations of various AI-related factors. The research results indicate that when multiple factors interact, elements such as the “human + AI” mixed labor force, digitization level, AI empowerment, and investments in AI capabilities collectively enhance the level of HRMVC. Therefore, under conditions of limited resources, prioritizing the use of AI tools for empowerment in HRMVC activities yields significant results. These results provide new insights into how AI can amplify traditional HRMVC factors, which aligns with studies suggesting the potential of AI in HRM [30]. Our research offers concrete evidence on how AI-related factors work together to significantly boost HRMVC, highlighting the importance of prioritizing AI investments in HRM practices, thus filling a critical gap in the literature.

4.1. Theoretical Contributions

This research proposed a comprehensive HRMVC influence factor model and conducted a simulation analysis based on this. The main theoretical contributions of this research are as follows:

(1)

The theoretical framework of HRMVC has been significantly extended: Combining grounded theory and system dynamics methods, this research not only identified the influencing factors of HRMVC but also explored the dynamic relationships among these factors, therefore expanding the theoretical framework of HRMVC. These include various environmental, organizational, and individual factors, such as AI empowerment and AI capability. These findings offer a more comprehensive understanding of how AI can be integrated into HRM processes, particularly in enhancing co-creation between different stakeholders. This addresses ongoing debates in the literature regarding the role of AI in transforming traditional HR practices, providing empirical support for AI’s potential to fundamentally reshape how value is co-created in HRM.

(2)

Advancing the application of value co-creation theory in HRM: By introducing the theory of value co-creation into enterprise HRM, this research reveals the interactive mechanisms among enterprises, government, and participants, enriching the application scenarios of the value co-creation theory. By doing so, it expands the scope of value co-creation theory beyond traditional contexts, showing its relevance and applicability in the modern HRM landscape, particularly in the era of digital transformation and AI integration. This research demonstrates how AI can serve as an active participant in co-creation processes, which not only broadens the application of value co-creation theory but also provides a new perspective on the interaction between technology and human resources. This contributes to the broader academic discourse on the implications of AI in organizational settings, particularly concerning ethical, practical, and strategic considerations.

(3)

Contributing to ongoing debates on AI in HRM: Against the backdrop of the AI era, this research systematically analyses the impact of AI-related factors on HRMVC and proposes specific strategies for empowering HRMVC activities using AI tools. By exploring how AI-related factors such as AI empowerment and AI capability interact with other HRMVC elements, this research provides empirical evidence that supports the strategic use of AI in HRM. This not only enriches the understanding of AI’s role in HRM but also addresses key questions in current debates about the future of work and the digital transformation of organizational processes.

4.2. Practical Implications

Based on the analysis of the results from various simulations of factors affecting HRMVC using system dynamics, this research suggests that companies focus on the following aspects when organizing and implementing HRMVC:

(1)

Pay attention to government policy adjustments. According to the analysis results of this research, government policies related to HRM, such as insurance, maternity, and taxation, can specifically influence the design of enterprise HRM processes and specific policies. These policies directly affect HRM stakeholders and have a substantial impact on enterprise HRM activities. Therefore, companies need to closely monitor relevant government policy trends and work closely with HRM stakeholders to promptly adjust the content, norms, and processes of HRMVC activities, thus effectively organizing enterprise HRMVC activities. Given the potential financial and operational impacts, companies need to implement proactive measures to monitor and adapt to policy changes. For instance, by anticipating shifts in taxation or labor laws, companies can optimize their HRMVC processes to maintain compliance while minimizing costs and avoiding potential disruptions to their operations.

(2)

Focus on providing support for enterprise HRMVC. While creating a fair, open, and innovation-supportive organizational atmosphere and corporate culture conducive to HRMVC, it is crucial to fully consider stakeholders’ opinions in the design of the HRM system and actively collect feedback during the implementation process to continuously optimize the HRM system and policies, establishing a suitable HRM framework for HRMVC activities. Additionally, it is necessary to build a resource platform for HRMVC, coordinating the diverse needs and relationships of participants, providing support from management, and keeping pace with technological advancements using AI to empower HRMVC activities. Moreover, to ensure economic efficiency and competitive advantage, companies should leverage AI to enhance decision-making processes and automate routine tasks within HRMVC. By building a robust resource platform that integrates AI technologies, companies can more effectively coordinate the diverse needs and relationships of participants, therefore improving overall productivity and responsiveness to market changes.

(3)

Enhance the effectiveness of HRMVC implementation from the perspective of HRMVC participants. Based on the different viewpoints and individual characteristics of the participants, it is important to enhance their identification with their own careers and the organization. Additionally, establish participants’ understanding of the future economy and company expectations and assist them in creating plans. To maximize the economic contributions of HRMVC, companies should invest in the development of participants’ professional, managerial, and AI skills, which not only improves HRMVC effectiveness but also ensures that the workforce is better equipped to drive innovation and respond to evolving market conditions. By doing so, companies can enhance both their internal capabilities and external market position.

(4)

In the new era of enterprise management, where AI is increasingly involved in enterprise HRMVC activities, companies must pay more attention to the impact of AI-related factors on HRMVC processes, such as “human + AI” mixed labor force, digitization level, AI empowerment, and investments in AI capabilities. When allocating resources, companies should prioritize empowering participants through AI in the design and implementation of HRMVC processes. By strategically increasing the use of ‘human + AI’ hybrid workforces and integrating AI as a core component of HRMVC, companies can enhance their operational efficiency and innovation potential. This approach not only drives cost reduction but also enables companies to rapidly adapt to market changes, therefore maintaining a competitive edge in a dynamic economic environment. Additionally, the use of “human + AI” hybrid workforces should be increased, with AI being introduced as a new type of workforce to participate in value creation. Companies should also focus on improving their level of digital intelligence to enhance their HRMVC capabilities. Furthermore, companies need to help participants fully understand the trend and efficiency of AI involvement in work. By enhancing their AI capabilities through learning, they can improve human-AI collaboration efficiency, which in turn can enhance the company’s HRMVC level.

4.3. Limitations and Future Directions

This research employs grounded theory and system dynamics simulation as research methods to analyze the factors of HRMVC based on semi-structured and secondary data and further research the influencing mechanism model. However, due to the limitations in sample size and research methodology, this research has the following shortcomings that need to be addressed in future research:

(1)

This research systematically explores the impact of environmental factors, organizational factors, and participant factors on HRMVC, providing a comprehensive framework. Future research could focus on a specific aspect or factor to delve deeper into the underlying mechanisms.

(2)

We observed that respondents with different demographic characteristics and personality traits exhibited varying degrees of corporate identification, execution feedback, and willingness to cooperate during the interviews. Subsequent research could further explore the boundaries of applicability for the theoretical model of influencing factors proposed in this research.

(3)

The research primarily considers the resource allocation combination after the integration of AI in the multi-factor simulation scheme. It does not cover all influencing factors, and future researchers can explore the impact mechanisms of other factor combinations on HRMVC.

(4)

This research did not address the implementation costs of different scenarios. Future research could further explore the cost-benefit analysis of HRMVC implementation plans to provide more comprehensive decision-making support.