How Does Industrial Intelligence Impact the Integration of the Industrial and Innovation Chains: Evidence from China

Abstract

Promoting the integration of the industrial and innovation chains (ICIC) constitutes a crucial strategy adopted by the Chinese government to foster sustainable economic development. Industrial intelligence (II), as a prominent application of artificial intelligence in the manufacturing sector, serves as a key engine for China’s industrial upgrading and has garnered widespread scholarly attention regarding its economic impacts. Using provincial-level panel data from China spanning 2011 to 2023, this study empirically investigates the impact of II on ICIC. The empirical results indicate the following: First, II exerts a significant positive impact on ICIC, and this conclusion remains robust after a series of robustness tests. Second, high-tech enterprises agglomeration and high-skilled labor agglomeration act as two critical channels through which II promotes ICIC, whereas technological innovation fails to play a mediating role. Third, both digital infrastructure and marketization positively moderate the relationship between II and ICIC, thereby significantly amplifying the positive impact of II on ICIC. Fourth, the positive effect of II on ICIC is found to be universally applicable: II can significantly promote ICIC in provinces with either strong or weak manufacturing (service) industries. These findings offer valuable theoretical support and practical implications for countries worldwide with diverse endowments in manufacturing and service industries that are pursuing II and striving to promote ICIC.

1. Introduction

Global warming has emerged as an imminent challenge that requires concerted efforts from all nations. Implementing strategies to reduce carbon emissions has become an important topic of discussion [1], and it is closely linked to broader discourses on digital transformation, industrial upgrading, and sustainable development. Addressing environmental problems poses a significant challenge for China’s manufacturing sector, especially during its current critical period of economic transformation—a context that is not unique to China. Manufacturing enterprises in China are confronted with the dual mandate of improving productivity while reducing environmental pollution [2], a dilemma that resonates with global discussions on balancing industrial growth and ecological sustainability in the era of intelligent-enabled production systems. Moreover, as one of the countries with the largest elderly populations, China is significantly affected by demographic changes, which further underscores the urgency of industrial structure upgrading [3]—a pressing issue also debated globally in the context of demographic transitions and industrial revitalization. Therefore, to pursue sustainable economic growth and improve citizens’ quality of life, China is actively promoting industrial transformation and fostering an industrial innovation ecosystem to explore new pathways for economic development [4], and its practices offer potential insights for other economies facing similar challenges of transformation, environmental governance, and demographic shifts.

According to the World Robotics Report 2024 published by the International Federation of Robotics (IFR), the global operational stock of industrial robots surpassed 4.2816 million units in 2024, with China accounting for 41% of the global market share. Industrial intelligence has emerged as a crucial driving force for environmental governance and economic sustainability in China. The existing literature demonstrates that industrial intelligence can significantly reduce energy intensity and improve corporate energy and resource utilization efficiency [5]. Meanwhile, the application of industrial intelligent technologies mitigates air pollution and lowers carbon emission intensity through two pathways: the improvement of energy utilization efficiency and the advancement of green technological innovation [6]. Transnational empirical evidence indicates that in the short run, industrial expansion driven by industrial intelligence leads to rising power consumption, thereby generating a certain carbon emission growth effect [7]. Nevertheless, in the long term, industrial intelligence accelerates the green transformation of industrial structure and reduces the over-reliance of economic growth on energy consumption, thereby curbing carbon intensity [8]. An empirical study by Song et al. (2025) [9] using data from 32 countries shows that industrial intelligence (II) has become a key driver for improving energy efficiency, cutting carbon emissions, and boosting economic growth. From the perspective of economic sustainability, industrial intelligence effectively improves total factor productivity and corporate labor productivity [10]. Even amid the shocks of economic policy uncertainty, it exerts a buffering effect on stabilizing production efficiency and optimizing factor allocation, consolidating the foundation for the sustainable development of the real economy [11] and ultimately achieving the sustainable development goals of ecological improvement and production efficiency enhancement [12]. An empirical study by Li et al. (2026) [13] using data from 53 countries worldwide shows that II can enhance a country’s global value chain resilience through three channels: labor substitution, trade cost reduction, and technological innovation, thereby promoting sustainable economic development.

There are numerous challenges in the construction of China’s innovation ecosystem and the realization of sustainable economic development. ICIC refers to the transformation of theoretical achievements into products through industry–technology integration, as well as the industrialization and scaling-up of products based on nodes in the industrial chain [14] ICIC can facilitate the precise matching of innovation resources with industrial demands, accelerate the transformation of basic research achievements into real productive forces, and enhance technological self-reliance in key links of the industrial chain [15]. It can improve the risk resilience and core competitiveness of the industrial chain, providing support for responding to external pressures. Thus, promoting ICIC is a strategic approach for China to construct an innovation ecosystem.

On the one hand, practical experience has shown that ICIC plays a significant role in advancing carbon neutrality in China [16]. For instance, enterprises in the innovation chain continuously yield green and digital technological achievements, while those in the industrial chain provide practical application scenarios for such technologies. A large number of cross-chain enterprise collaborations can accelerate the industrialization and large-scale application of diverse green and low-carbon technologies, facilitate the phase-out of high-energy-consuming and inefficient production capacity, and promote industrial upgrading toward high-end and low-carbon development, thereby achieving the coordinated development of economic growth and ecological environmental improvement. On the other hand, against the backdrop of increasingly fierce China–U.S. strategic competition, the security and stability of China’s industrial chain are facing severe challenges, which stem from the vulnerability of the industrial chain caused by external technological blockades and industrial containment. The United States has introduced a series of policies, such as the CHIPS and Science Act, the National Quantum Initiative Act, and the America COMPETES Act of 2022. By restricting the export of high-end technologies and erecting technological barriers, these policies have disrupted China’s connections with the global innovation chain and industrial chain, and hindered China’s industrial upgrading and technological development. In this context, industrial innovation in China has become a core issue in safeguarding national economic security. Promoting ICIC can bridge the links between basic research, technological breakthroughs and industrial transformation. Meanwhile, it can empower high-end industrial upgrading, liberate industries from low-end lock-in in the global value chain, and effectively tackle the dilemma of external decoupling and supply chain disruption.

However, in practice, China’s ICIC still faces many challenges. Specifically, there is an imbalance between China’s investment in basic research and the efficiency of achievement transformation. Although investment in basic research by universities and research institutes continues to increase, the channel for transforming scientific and technological achievements from laboratories to industrial markets remains blocked. The pilot test service system and the mechanism for the transformation of scientific and technological achievements are still inadequate, resulting in a large number of innovative achievements remaining at the sample stage and failing to effectively match the needs of industrial chain upgrading.

Some literature has measured the ICIC level in China using various methods [15,17,18,19]. Overall, the ICIC level is high in eastern China but relatively low in western China. Previous studies have found that factor endowments [20], institutional environment [21], human capital [22], demographic structure [23], green credit [24], technological innovation [25], digital economy [26], and industrial intelligence [27] can promote the development of the industrial chain. Meanwhile, innovation climate [28], innovation policy [29], supply chain learning [30], and supply chain collaboration [31,32] are important factors driving the development of innovation chain. Currently, some studies have explored the influencing factors of ICIC and verified the positive impacts of economic benefits [33], blockchain technology [34], global value chain (GVC) participation [35], human resources, capital, and information [14] on ICIC.

II represents a critical embodiment of technological progress. By deeply integrating digital technologies and intelligent equipment with industrial scenarios, it can drive business model innovation [36], break down barriers between the industrial chain and innovation chain, and facilitate the precise alignment of innovation resources with industrial demands, thereby reshaping the innovation ecosystem [37]. Recent research on II has primarily focused on its effects on various factors, including energy intensity [5], air pollution [6], carbon emissions [7], carbon intensity [8], total factor productivity [10], labor productivity [11], energy efficiency [12], and others. However, few studies have explored the relationship between II and ICIC, and quantitative research in this field is almost non-existent. A recent study shows that IoT-enabled industrial intelligence helps improve the integration efficiency between the industrial chain and the innovation chain in high-tech manufacturing [38]. Li et al. (2025) [39] support this view by arguing that the Internet of Things and blockchain technologies play an important role in promoting ICIC. They conduct an in-depth technical analysis of how intelligent technologies enhance firms’ operational efficiency and innovation efficiency. However, none of these studies empirically examine how II impacts ICIC.

This paper makes three key contributions. First, against the dual constraints of China’s green development and demographic transition, as well as the realistic demand for industrial modernization, this paper explores the environmental and economic effects of II, as well as the significance of ICIC for the sustainable development of the environment and economy. It breaks the research limitation of prior studies that treat II and ICIC as two separate research domains, and further enriches the empirical research system of II and ICIC in the artificial intelligence era. Second, this paper systematically explores the internal mechanism and transmission channels of how II affects ICIC. In the context of intensifying China–U.S. strategic competition and ongoing external technological blockades and industrial containment, it provides a novel analytical perspective and solid theoretical support for breaking the development bottlenecks of ICIC, optimizing resource allocation between industrial and innovation chains, and enhancing the independent controllability and anti-risk capacity of China’s industrial chain. This study effectively fills the research gap in the existing literature, which lacks in-depth discussion on the intrinsic logical relationship and mediating pathways between II and ICIC. Third, this paper provides targeted practical implications and policy references for China’s manufacturing sector to simultaneously achieve carbon emission reduction, productivity improvement, and industrial chain security, as well as to optimize and upgrade the national industrial innovation ecosystem. Furthermore, the research conclusions and policy recommendations are also universally instructive and applicable for other economies with different manufacturing and service industry foundations, offering valuable lessons for promoting II and ICIC worldwide.

2. Literature Review and Hypothesis Development

2.1. The Impact of II on ICIC

II refers to the intelligent upgrading of traditional manufacturing through the application of artificial intelligence (AI) technologies. Relying on the unique characteristics of AI—such as permeability, substitutability, synergy, and creativity [40]—and its important role in reshaping industrial development and optimizing industrial structure, II can comprehensively promote ICIC from three dimensions.

First, relying on the permeability and synergy of AI technology, II can break down information barriers between the industrial chain and innovation chain [41]. This transformation enables II to closely link the R&D segment of the innovation chain (e.g., universities and research institutions) with the production and operation segment of the industrial chain (e.g., enterprises) through intelligent technologies, thereby achieving real-time alignment between innovation needs and industrial demands. Specifically, enterprises can timely transmit technical bottlenecks and market demands encountered in production to R&D entities via intelligent platforms. Supported by intelligent production and management systems, R&D achievements can be rapidly tested, transformed, and applied in industrial production. This process not only fully reflects the optimizing effect of AI technology on enterprise R&D and production processes [42], but also effectively shortens the transformation cycle of innovative achievements, strengthens the connection and interaction between the two chains, and lays a solid foundation for their integration. More importantly, this integration mechanism helps allocate innovative and industrial resources in an economical and low-carbon manner, reduces redundant R&D and inefficient production, and forms an endogenous driving force for the sustainable operation and long-term evolution of industrial and innovation systems.

Second, the creativity and substitutability of AI technology drive II to continuously promote industrial upgrading, which in turn provides a favorable industrial foundation and resource guarantee for ICIC. II can improve enterprise productivity, boost the development of high-tech manufacturing and high-tech services, and drive industries to transform from low value-added to high value-added activities [43], thereby guiding the innovation chain to focus on high-end technology research and development and realize the simultaneous upgrading of the two chains. An empirical study by Unilu & Kocak (2026) [44] using data from 60 countries worldwide shows that II can enhance a country’s green export competitiveness through improved technological efficiency and sustainability-oriented innovation. By promoting the deep integration of AI and big data technologies, II can facilitate the cross-industry and cross-regional flow of production factors, optimize the allocation of industrial resources [45], and promote the coordinated development of related industries within the industrial chain. Meanwhile, it encourages the innovation chain to carry out cross-field and cross-disciplinary collaborative innovation, breaks the fragmented development pattern of the industrial chain and innovation chain, improves the coordination and integration efficiency of the whole system, and ultimately provides stable industrial ecological support for ICIC. From a sustainability perspective, II-driven industrial upgrading can accelerate the phase-out of high-energy-consumption and high-pollution backward capacity, promote the penetration of green production technologies and low-carbon innovation achievements, and help balance industrial economic growth with carbon emission reduction.

Third, II can reshape enterprise productivity and drive the reform of social production relations [46], thereby optimizing the micro-foundation and institutional environment for ICIC. On the one hand, II can optimize enterprises’ R&D, production and management processes [42], enhance their ability to absorb and transform innovative achievements, and make enterprises the core carrier connecting the industrial chain and innovation chain, thus consolidating the micro-subject foundation for ICIC. On the other hand, II can promote the emergence of many emerging industries [47], which not only enriches the content and extension of the industrial chain, but also generates new innovation demands and forces the continuous upgrading of the innovation chain. A long-term mechanism of mutual promotion and positive interaction between industrial expansion and innovation iteration is thus formed, which ultimately drives the integration of the industrial and innovation chains and injects sustained impetus into high-quality industrial development. Furthermore, the micro-level optimization of enterprise production modes and the emergence of intelligent green emerging industries can consolidate the long-term operational vitality of ICIC and realize the dynamic sustainability of industrial innovation, economic growth and ecological governance.

Accordingly, Hypothesis 1 is proposed.

Hypothesis 1.

Industrial intelligence helps to promote the integration of the industrial and innovation chains.

2.2. The Impact Channel of II on ICIC

The essence of II lies in the integration of AI technologies with industry-specific scenarios, thereby enabling intelligent decision-making and optimizing resource allocation in product design, production, operation, management, and other activities [48]. The realization of this function requires the development of new technologies such as big data and AI algorithms. Such technological advances are crucial for designing software systems that achieve human–machine collaborative management and intelligent decision-making. Therefore, it can be inferred that the demand for intelligent technologies driven by II will promote the agglomeration of technology-intensive producer services, including information services, R&D services, and technical services, especially in the fields of R&D, design, and information technology [49]. An empirical study by Algül (2024) [50] using data from 17 countries worldwide shows that the adoption of artificial intelligence technology significantly raises employment in the AI-related service sectors while reducing employment in the industrial and agricultural sectors. Paul et al. (2025) [51] hold a similar view, finding that the growth of industrial robots in India coincides with the stagnation of low-skilled labor, while high-skilled labor expands. The agglomeration of technology-intensive producer services will further drive the agglomeration of high-tech enterprises, thereby forming a coordinated development pattern between the service industry and manufacturing industry [52]. These high-tech enterprises cover a wide range of fields closely related to industrial intelligence, including artificial intelligence research and development, big data application, and intelligent equipment manufacturing. They will form close industrial linkages with manufacturing enterprises implementing II projects and various producer services, thereby strengthening the local innovation ecosystem. This ecosystem can serve as an important carrier for ICIC. Within this ecosystem, high-tech enterprises act as the core R&D entities of the innovation chain, focusing on breakthroughs in core technologies and the output of innovative achievements. Manufacturing enterprises, as the core production entities of the industrial chain, concentrate on the industrial application and large-scale production of innovative achievements. Relying on geographical proximity and industrial correlation formed by agglomeration, the two sides break down barriers between the industrial chain and innovation chain, enabling the precise matching of innovation resources with industrial demands.

By cooperating with manufacturing enterprises implementing II projects, these producer services and high-tech enterprises can promote technology spillovers to the manufacturing sector through spatial interactions of knowledge, information, and other factors, thereby increasing the value-added of local industries [53]. As a bridge between high-tech enterprises and manufacturing, producer services play an important role in promoting technological cooperation between the two sides [52]. On the one hand, high-tech enterprises rapidly transfer R&D achievements to manufacturing enterprises for industrial application through producer services, which shortens the transformation cycle of innovative achievements and strengthens the supporting role of the innovation chain for the industrial chain. On the other hand, through producer services, high-tech enterprises can accurately capture the technological needs of manufacturing enterprises in the process of II and thus carry out targeted R&D activities. This aligns the R&D direction of the innovation chain with the development needs of the industrial chain and promotes ICIC.

According to the capital-skill complementarity hypothesis, new capital goods induced by technological progress are significantly complementary with high-skilled labor, while they substitute for low-skilled labor [54]. In the context of II, the popularization of new capital goods such as intelligent equipment, data processing systems, and industrial software requires labor to possess professional capabilities including R&D and design, equipment operation and maintenance, data interpretation and analysis, and process optimization. These capabilities are exactly the core endowments of high-skilled labor [39]. II is not simply equipment substitution, but promotes continuous technological innovation and product upgrading in enterprises. High-skilled labor is the core entity of innovation activities, and the demand for high-skilled labor will increase with the improvement of enterprises’ innovation intensity [55]. In fact, high-skilled labor, which is compatible with new industrial forms such as high-end manufacturing and the integration of producer services, serves as the core human capital support in the integration of the industrial and innovation chains.

According to the theory of external economies, industrial agglomeration can generate three major externalities: labor pool sharing, intermediate input sharing, and knowledge spillovers [56]. First, cross-industry labor mobility and knowledge exchange are conducive to radical innovation [57], which can explain the positive effect of collaborative agglomeration between high-tech enterprises and producer services on urban innovation. Second, from the perspective of competition and collaboration, fierce market competition within industrial agglomerations compels enterprises to sustain continuous R&D investment, while frequent face-to-face interactions reduce cooperation and learning costs, jointly driving the upgrading of regional innovation capacity [58]. An empirical study finds that the agglomeration of high-tech enterprises and high-skilled labor helps promote technological innovation, foster the development of innovative cities, and accelerate economic transformation [59]. The mobility of high-skilled labor among these high-tech enterprises, producer service enterprises, and manufacturing enterprises facilitates technology absorption and productivity improvement, thereby advancing the development of urban innovation ecosystems [60]. More importantly, through collaborative innovation and resource sharing, these enterprises can continuously improve the R&D and transformation links of the innovation chain, optimize the production and operation links of the industrial chain, and further promote ICIC.

Accordingly, Hypotheses 2a–c are proposed.

Hypothesis 2a.

II promotes ICIC by facilitating the agglomeration of high-tech enterprises.

Hypothesis 2b.

II promotes ICIC by facilitating the agglomeration of high-skilled labor.

Hypothesis 2c.

II promotes ICIC by facilitating technological innovation.

2.3. The Moderating Effect of Digital Infrastructure and Marketization

Digital infrastructure is the fundamental carrier underpinning the application of II, exerting a critical impact on the transmission efficiency of industrial intelligence technologies and the smooth progress of ICIC. By breaking geographical barriers, digital infrastructure—such as industrial internet platforms, 5G networks, and big data centers—can enable real-time interconnection among the R&D segment of the innovation chain (universities and research institutions), the industrial intelligence application segment (manufacturing enterprises), and the production segment of the industrial chain (upstream and downstream firms) [61]. This can mitigate information asymmetry and transaction costs in the commercialization of innovative outcomes [62], facilitating the rapid penetration of II technologies across all links of the industrial chain and innovation chain, thereby reinforcing the positive effect of II on ICIC.

Furthermore, digital infrastructure can optimize pathways for knowledge spillovers, accelerate the conversion of R&D outputs into industrial productivity, and further promote ICIC. Specifically, by expediting cross-temporal and cross-spatial knowledge flows and enhancing collaborative innovation among stakeholders, digital infrastructure can transform geographically dispersed technical knowledge into analyzable, transferable, standardized information modules, significantly reducing the tacitness of technical knowledge [63]. Moreover, industrial internet platforms can accelerate the diffusion of technologies and knowledge along supply chains, shorten technology adoption lags, and expedite the translation of R&D achievements into the industrialization phase [64]. Conversely, inadequate development of digital infrastructure and insufficient digital connectivity will impede the effective application of II technologies, prolong the conversion cycle of innovative outcomes, and hinder the linkage between the industrial chain and innovation chain.

Marketization is a core indicator measuring regional resource allocation efficiency, market entity vitality, and the improvement of the institutional environment [65]. It can provide a sound institutional environment and market momentum for II to promote ICIC.

First, a higher level of marketization is associated with a sounder market competition mechanism and greater competitive pressure faced by enterprises. This compels enterprises to match innovation chain resources more efficiently, increase investment in II, promote technological upgrading and business model innovation, and accelerate the transformation of innovative achievements [66], thereby strengthening the promoting effect of II on ICIC. In addition, a sound market mechanism helps foster a favorable competitive environment, drives innovative resources to agglomerate toward efficient and innovative enterprises, and promotes collaboration between the manufacturing industry and producer services [67]. Second, a higher level of marketization implies less government intervention in economic activities, which can improve the efficiency of market-based factor allocation and facilitate the free flow of innovative factors such as talent, technology, and capital [68], thus promoting ICIC. Third, regions with a higher level of marketization tend to have a more complete intellectual property protection system and more standardized market order [69]. This can more effectively protect the legitimate rights and interests of innovation entities, reduce the risk of infringement on innovative achievements, and better stimulate the innovation enthusiasm of enterprises and research institutions [70]. In such an environment, enterprises can form broader cooperation networks and possess stronger technological absorptive capacity, which can further promote technological exchange and cooperation [71]. Conversely, a low level of marketization is characterized by insufficient market competition, inefficient factor allocation, and inadequate intellectual property protection. It will reduce enterprises’ innovation motivation, leave the synergy between the industrial chain and the innovation chain without sufficient market support, and thus weaken the promoting effect of II on ICIC.

Accordingly, Hypotheses 3a,b are proposed.

Hypothesis 3a.

Digital infrastructure strengthens the positive effect of II on ICIC.

Hypothesis 3b.

Marketization strengthens the positive effect of II on ICIC.

3. Research Design

3.1. Econometric Model

The following regression model is constructed to examine the impact of II on ICIC.

$$ICI{C}_{it}={\beta }_0+{\beta }_{1}I{I}_{it}+{\displaystyle \sum {\beta }_{i}CONTRO{L}_{it}}+{\mu }_i+{\delta }_t+{\epsilon }_{it}$$

(1)

where subscript i and t denote the province and year, ICIC_it denotes integration of the industrial and innovation chains, II_it denotes industrial intelligence, CONTROL_it denotes control variables. μ_i and δ_t denote the province and year fixed effect, while ε_it denotes the error term.

3.2. Variable Measurement and Description

3.2.1. Dependent Variable

Integration of the industrial and innovation chains (ICIC) is the dependent variable. Drawing on Wang & Zhao (2026) [72], this paper uses the coupling coordination degree model to measure the level of ICIC. The coupling coordination degree mainly reflects the closeness and coordination between the industrial chain and the innovation chain, as well as their interactive and synergistic relationship during development. The coupling coordination degree model is theoretically appropriate for measuring ICIC, as it is specifically designed to capture the interactive intensity and synergistic matching between two interdependent subsystems, which is highly consistent with the core connotation of ICIC. This model can comprehensively reflect both the developmental levels of the two chains and their coordinated evolution, thus ensuring the rationality and validity of the measurement. Drawing on Hu & Zhang (2023) [15], Wang & Zhao (2023) [18], and Wang & Zhao (2026) [72], we quantify the indicators related to the industrial chain and the innovation chain separately. For the industrial chain subsystem U₁, evaluation indicators are selected from three dimensions: industrial foundation, industrial economy, and industrial linkage. For the innovation chain subsystem U₂, evaluation indicators are selected from three dimensions: innovation input, innovation output, and innovation environment.

$$C=2\sqrt{{\displaystyle \frac{U_1\cdot U_2}{{(U_1+U_2)}^2}}}$$

(2)

$$T=a\cdot U_1+b\cdot U_2$$

(3)

$$ICIC=D=\sqrt{C\cdot T}$$

(4)

where C is the coupling degree between the industrial chain and the innovation chain. U₁ is the evaluation index of the industrial chain subsystem, and U₂ is the evaluation index of the innovation chain subsystem. T is the comprehensive coordination level of the industrial chain subsystem and the innovation chain subsystem. a and b denote the weights of the industrial chain evaluation index and the innovation chain evaluation index, respectively, reflecting the contribution of each subsystem to the comprehensive coordination level, where a = b = 0.5. D is the coupling coordination degree of the industrial chain and innovation chain system, where D = ICIC. The higher the value of D, the better the ICIC.

To objectively reflect the importance of each indicator in the overall evaluation, the original data are first standardized. Then, the entropy weight method is used to assign weights to the 16 secondary indicators of the industrial chain and innovation chain subsystems across 30 Chinese provinces (see

Table 1. ICIC evaluation index system.

Tier 1 Indicators	Tier 2 Indicators	Tier 3 Indicators	Measurement	Weight
Industrial chain subsystem (U1)	Industrial Foundation	Main Business Revenue	Main business revenue of industrial enterprises above designated size	X1 = 0.0820
		Total Assets	Total assets of industrial enterprises above designated size	X2 = 0.0636
		Total Profits	Total profits of industrial enterprises above designated size	X3 = 0.0482
		Number of Enterprises	Number of industrial enterprises above designated size	X4 = 0.0915
		Urban Employment	Number of urban employed persons	X5 = 0.0447
	Industrial Economy	Industrial Profitability	Main business revenue/Main business cost of industrial enterprises above designated size	X6 = 0.4823
		Degree of Economic Servitization	Ratio of added value of the tertiary industry to that of the secondary industry	X7 = 0.0616
	Industrial Linkage	Proportion of Foreign-invested Enterprises	Number of foreign-invested enterprises/Number of industrial enterprises above designated size	X8 = 0.1098
		Employment Contribution Rate	Urban employment in manufacturing industry/Total urban employed persons	X9 = 0.0163
Innovation chain subsystem (U2)	Innovation Input	R&D Personnel Input	Full-time equivalent of R&D personnel in industrial enterprises above designated size	X10 = 0.1203
		R&D Expenditure	R&D expenditure of industrial enterprises above designated size	X11 = 0.1445
	Innovation Output	Intellectual Property Acquisition Rate	Number of patent applications of industrial enterprises above designated size	X12 = 0.1794
		Proportion of New Product Revenue	Sales revenue of new products of industrial enterprises above designated size	X13 = 0.1889
	Innovation Environment	Government Support	Science and technology expenditure/General public budget expenditure	X14 = 0.0679
		Innovation Activity	Turnover of technology market	X15 = 0.2609
		University Participation	Number of institutions of higher education	X16 = 0.0382

). The entropy weight method determines weights based on the information entropy and dispersion degree of each indicator, which enables an objective identification of the information content contained in different indicators. Therefore, the employment of the entropy weight method ensures the rationality and reliability of the constructed ICIC index.

3.2.2. Independent Variable

Industrial intelligence (II) is the independent variable. Drawing on Zhang & Zhu (2023) [7], the installation density of industrial robots is employed to quantify the level of II in each province of China. The formula is as follows:

$$I{I}_{it}={\displaystyle \frac{{\displaystyle {\sum }_{j=1}^{14}(\frac{Z_{itj}}{Z_{tj}}\cdot R_{tj})}}{LABO{R}_{it}}}$$

(5)

where II_it denotes industrial intelligence in period t of province i, with j = 1, 2, 3, …, 14 denoting the 14 manufacturing industries in China. Z_itj denotes the number of employees in industry j in period t of province i, while Z_tj represents the total number of employees in industry j in period t across China. R_tj denotes the total number of industrial robots installed in industry j in period t across China, and LABOR_tj represents the total number of manufacturing employees in period t of province i.

3.2.3. Mediating Variables

High-tech enterprise agglomeration is the first mediating variable. Drawing on Han (2023) [73], the location entropy index is employed to assess the level of high-tech industry agglomeration in each province of China. The formula is as follows:

$$HT{A}_{it}={\displaystyle \frac{H{T}_{it}/T_{it}}{H{T}_t/T_t}}$$

(6)

where HTA_it denotes high-tech enterprises agglomeration in period t of province i. HT_it denotes the number of high-tech enterprises in period t of province i, while HT_t denotes the number of high-tech enterprises in period t across China. T_it denotes the total number of enterprises in period t of province i, and T_t denotes the total number of enterprises in period t across China.

High-skilled labor agglomeration is the second mediating variable. The location entropy index is employed to assess the level of high-skilled labor agglomeration in each province of China. The formula is as follows:

$$HS{A}_{it}={\displaystyle \frac{H{S}_{it}/S_{it}}{H{S}_t/S_t}}$$

(7)

where HSA_it denotes high-skilled labor agglomeration in period t of province i. HS_it denotes the number of high-skilled labor in period t of province i, while HS_t denotes the number of high-skilled labor in period t across China. S_it denotes the total number of labor in period t of province i, and S_t denotes the total number of labor in period t across China.

Technological innovation is the third mediating variable. Drawing on Zhang et al. (2024) [74], the number of invention patent applications filed in a province in the current year is adopted to measure the level of technological innovation.

3.2.4. Moderating Variables

Digital infrastructure (DI) is the first moderating variable. Drawing on Liang & Liu (2026) [75] and Gao et al. (2026) [76], we construct a comprehensive evaluation index for the digital infrastructure level of each province from four dimensions: internet penetration rate (number of broadband internet access users per 100 people), the scale of relevant practitioners (the proportion of practitioners in computer services and software industry in urban unit practitioners), application scale (per capita total telecommunication business volume), and mobile phone penetration rate (number of mobile phone users per 100 people). The weights of each indicator are determined by the entropy weight method.

Marketization (MARK) is the second moderating variable. Drawing on Ren et al. (2024) [77], the marketization index from China’s Provincial Marketization Index Report is adopted to measure the marketization level of each province. This index comprehensively takes into account the relationship between the government and the market, the development of the non-state-owned economy, the development level of the product market, the development level of the factor market, as well as the development of market intermediary organizations and the legal environment.

3.2.5. Control Variables

ICIC may also influenced by various factors, such as fiscal expenditure (FE), financial development (FD), economic development (GDP), foreign direct investment (FDI), investment in the information industry (INVEST), environmental regulation (ER), and transportation infrastructure (TRANS).

Table 2. Variable definitions.

Variable	Name	Calculation/Value
ICIC	Integration of industrial chain and innovation chain	measured by the coupling coordination degree model, see Equations (2)–(4) for details.
II	Industrial intelligence	Measured by installation density of industrial robot, see Equation (5) for details.
HTA	High-tech enterprise agglomeration	Measured by location entropy index, see Equation (6) for details.
HSA	High-skilled labor agglomeration	Measured by location entropy index, see Equation (7) for details.
TI	Technological innovation	TI = ln(the number of invention patent applications filed in the province in the current year + 1)
DI	Digital infrastructure	Measured by the entropy weight method.
MARK	Marketization	the overall marketization level of each province, derived from the China Provincial Marketization Index Report
FE	Fiscal expenditure	FE = Fiscal expenditure/GDP
FD	Financial development	FD = Deposit and Loan Balance/GDP
GDP	Regional economic development	GDP = GDP growth rate
FDI	Foreign direct investment	FDI = FDI/GDP
INVEST	Investment of information industry	INVEST = Fixed asset investment in information transmission software and information technology service industry/total investment in fixed assets
ER	Environmental regulation	ER = Industrial pollution control investment/GDP
TRANS	Transportation infrastructure	TRANS = Number of public buses per 10,000 people

shows the variable definitions.

3.3. Data Sources

The raw data for measuring ICIC are from the China Statistical Yearbook; industrial robot data (the raw data for measuring II) are from the IFR (International Federation of Robotics) database; labor-related data are from the China Labor Statistical Yearbook; high-tech enterprise and high-skilled labor data are from the China Torch Statistical Yearbook; the raw data for measuring DI are from the China City Statistical Yearbook; MARK data are from the Provincial Marketization Index Report of China; and all other data are from the China Statistical Yearbook.

4. Results

4.1. Descriptive Statistics

Table 3. Descriptive statistics for all variables.

Variable	Obs	Mean	D.	Min	Max
ICIC	390	0.265	0.105	0.092	0.535
II	390	0.391	0.503	0.007	2.234
HTA	390	0.775	0.450	0.130	2.182
HSA	390	0.794	0.581	0.202	3.144
TI	390	9.054	1.567	5.081	12.139
DI	390	0.242	0.182	0.058	0.991
MARK	390	8.017	1.864	3.564	11.494
FE	390	8.033	0.744	5.783	9.506
FD	390	28.095	1.003	25.342	30.116
GDP	390	0.123	0.075	−0.168	0.271
FDI	390	23.739	1.821	18.020	26.045
INVEST	390	0.012	0.008	0.001	0.042
ER	390	21.074	0.954	17.952	23.149
TRANS	390	12.094	3.254	6.850	24.670

reports the descriptive statistics. The minimum, maximum, and mean values of ICIC are 0.092, 0.535, and 0.265, respectively, with a standard deviation of 0.105. This indicates that ICIC in China is relatively low, leaving considerable room for improvement, and that there exist significant disparities across provinces. The minimum, maximum, and mean values of II are 0.007, 0.547, and 0.392, respectively, with a standard deviation of 0.503. There are two reasons for large standard deviation. First, significant regional heterogeneity exists across Chinese provinces: industrial intelligence is high in the eastern regions but much lower in the western regions. Second, rapid and uneven growth since 2011 has further widened cross-provincial differences in robot adoption. Overall, the level of industrial intelligence in China is still low on average, and some provinces have barely installed industrial robots, indicating substantial imbalances in intelligent transformation.

4.2. Baseline Regression Results

In Models (1), (2), and (3) of

Table 4. Baseline regression results.

Variable	(1)	(2)	(3)
	ICIC	ICIC	ICIC
II	0.030 ***	0.023 ***	0.027 ***
	(0.006)	(0.005)	(0.005)
FE		0.008	0.009
		(0.016)	(0.016)
FD		−0.037 ***	−0.038 ***
		(0.009)	(0.009)
GDP		0.107 ***	0.111 ***
		(0.024)	(0.025)
FDI		0.006 ***	0.006 ***
		(0.002)	(0.002)
INVEST			0.284
			(0.272)
ER			−0.001
			(0.002)
TRANS			0.002 ***
			(0.001)
_CONS	0.253 ***	1.085 ***	1.081 ***
	(0.002)	(0.214)	(0.212)
PROVINCE	YES	YES	YES
YEAR	YES	YES	YES
N	390	390	390
R2	0.082	0.232	0.254
F	25.756	27.346	20.115
p	0.000	0.000	0.000

Note: Robust standard errors are in parentheses, *** indicate significant levels at 1%.

, the regression coefficients of II are all significantly positive. These findings indicate that II can promote ICIC. The potential reasons are as follows: (1) II can shorten the cycle of achievement transformation by breaking information barriers and connecting innovation and industrial links, enabling innovative achievements to quickly penetrate into industrial production and thus directly promoting ICIC. II can integrate production, R&D, and demand data to enable targeted R&D, link upstream innovation with downstream applications, and accelerate testing and deployment through smart equipment, thus greatly shortening commercialization cycles and advancing the integration of innovation and industrial chains. (2) II can provide an industrial foundation and ecological support for the integration of the two chains by driving industrial upgrading. On the one hand, II can promote the intelligent transformation of production links, improve industrial standardization, digitization and intensification, and form a modern industrial system that is more compatible with advanced technological achievements. On the other hand, it can foster a collaborative ecology including R&D, production, application and iteration, enabling technological innovation and industrial development to promote each other. This upgraded industrial foundation and sound ecology can reduce the cost of achievement docking and application, stabilize the transformation path, and provide sustained support for the deep integration of the two chains. (3) II can strengthen the long-term mechanism for the integration of the two chains by consolidating microeconomic entities and fostering emerging industries. It can facilitate the emergence of industries related to intelligent manufacturing, which act as a bridge between technological breakthroughs and industrial applications. The joint effect of stable micro-entities and booming emerging industries helps form a self-reinforcing cycle of innovation and industrial development, thus supporting the long-term and in-depth integration of the two chains. In addition, II can not only improve production efficiency but also create growth opportunities for producer services in intelligence-related technology fields. This, in turn, promotes the synergistic development of the secondary and tertiary industries and ultimately promotes ICIC.

4.3. Robustness Test

To verify the robustness of the baseline regression results, we conduct a series of robustness tests. First, we alter the measurement method of the independent variable II. The ratio of total industrial robot installations to total urban non-private sector employees is employed as an alternative measure of the level of II in each province. The regression results are reported in Model (1) of

Table 5. Robustness regression results.

Variable	(1)	(2)	(3)	(4)
	ICIC	ICIC	ICIC	ICIC
II	0.100 ***	0.028 ***	0.035 ***	0.027 ***
	(0.019)	(0.007)	(0.006)	(0.005)
FE	0.011	0.018	0.054 ***	0.008
	(0.016)	(0.015)	(0.008)	(0.016)
FD	−0.037 ***	−0.034 ***	−0.007	−0.037 ***
	(0.008)	(0.008)	(0.007)	(0.009)
GDP	0.113 ***	0.104 ***	0.094 ***	0.103 ***
	(0.024)	(0.026)	(0.032)	(0.026)
FDI	0.006 ***	0.006 ***	0.002	0.006 ***
	(0.002)	(0.002)	(0.002)	(0.002)
INVEST	0.396	−0.039	0.337	0.302
	(0.271)	(0.273)	(0.243)	(0.276)
ER	−0.002	−0.000	0.000	−0.001
	(0.002)	(0.002)	(0.002)	(0.002)
TRANS	0.002 **	0.001 *	0.002 **	0.002 **
	(0.001)	(0.001)	(0.001)	(0.001)
_CONS	1.056 ***	0.895 ***	−0.082	1.058 ***
	(0.208)	(0.219)	(0.176)	(0.216)
PROVINCE	YES	YES	YES	YES
YEAR	YES	YES	YES	YES
N	390	360	390	360
R2	0.286	0.249	0.316	0.248
F	18.948	18.118	30.037	18.608
p	0.000	0.000	0.000	0.000

Note: Robust standard errors are in parentheses, ***, **, and * indicate significant levels at 1%, 5%, and 10%, respectively.

. Second, considering the potential lagged effect of II on ICIC, the independent variable II is lagged by one period. The regression results are reported in Model (2) of Table 5. Third, to address the impact of extreme values, all continuous variables are winsorized at the 5% level. The regression results are reported in Model (3) of Table 5. Finally, to eliminate the impact of the COVID-19 pandemic, observations from 2020 are excluded from the sample. The regression results are reported in Model (4) of Table 5. In Table 5, the regression coefficients of II are all significantly positive, consistent with the baseline regression results. This indicates that the baseline regression findings of this paper are robust.

4.4. Endogeneity: 2SLS Estimation

The integration of the industrial and innovation chains may incentivize local manufacturing enterprises to accelerate industrial intelligence transformation. If a bidirectional causal relationship exists between II and ICIC, endogeneity problems may arise. In addition, this study may omit some important explanatory variables, which may also lead to endogeneity. Therefore, this paper divides China into three major economic regions: Eastern, Central, and Western China. The mean value of industrial robot installation density in other provinces within the same economic region (excluding the focal province) in the same period is used as the instrumental variable for industrial robot application (II). This IV satisfies the relevance condition because industrial robot adoption within the same economic region is shaped by regional factor endowments, industrial layouts, infrastructure conditions, and technology diffusion trends. As a result, the IV is strongly correlated with industrial robot usage in the focal province. Meanwhile, the IV is exogenous and satisfies the exclusion restriction. Since the IV excludes the focal province itself, it is primarily determined by economic conditions and technology adoption behaviors of other provinces, rather than by province-specific unobserved shocks that simultaneously affect II and ICIC. Although regional policies or macroeconomic shocks may jointly influence II and ICIC across provinces, the empirical model in this study already controls for year fixed effects and province fixed effects, which absorb nationwide and time-invariant provincial shocks. The above design renders the instrumental variable theoretically plausible and empirically valid, satisfying both the relevance and exclusion restrictions to a large extent. Therefore, the instrumental variable approach adopted in this study effectively mitigates endogeneity concerns and lends credible support to the causal interpretation of the empirical results. The 2SLS regression results are reported in

Table 6. 2SLS regression results.

Variable	(1)	(2)
	II	ICIC
	First Stage	Second Stage
II_MEAN	0.563 ***
	(0.165)
II		0.039 ***
		(0.015)
CONTROL	YES	YES
PROVINCE	YES	YES
YEAR	YES	YES
N	390
Kleibergen–Paap rk LM statistic	12.38 (0.0004)
Cragg–Donald Wald F statistic	38.39

Note: Robust standard errors are in parentheses, *** indicate significant levels at 1%.

.

In Table 6, p-value associated with the first-stage KP-LM statistic is 0.0004, providing strong evidence against the Unidentifiable Assumptions. The CD-Wald F-value is 38.39, greater than 16.38, indicating that there is no weak instrument problem.

5. Further Analysis

5.1. The Impact Channels of II on ICIC

To explore the mechanisms through which II promotes ICIC, the following regression model are conducted.

$$MEDIATIN{G}_{it}={\beta }_0+{\beta }_{1}I{I}_{it}+{\displaystyle \sum {\beta }_{i}CONRO{L}_{it}}+{\lambda }_i+{\delta }_t+{\epsilon }_{it}$$

(8)

$$ICI{C}_{it}={\phi }_0+{\phi }_{1}I{I}_{it}+{\phi }_{2}MEDIATIN{G}_{it}+{\displaystyle \sum {\phi }_{i}CONTRO{L}_{it}}+{\lambda }_i+{\delta }_t+{\epsilon }_{it}$$

(9)

MEDIATING_it denotes the mediating variables, which are high-tech enterprise agglomeration (HTA), high-skilled labor agglomeration (HSA), and technological innovation (TI), respectively.

The regression results are shown in Models (1)–(6) of

Table 7. Impact channels of II on ICIC.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	HTA	ICIC	HSA	ICIC	TI	ICIC
HTA		0.017 ***
		(0.006)
HSA				0.016 ***
				(0.003)
TI						0.011 ***
						(0.003)
II	0.107 **	0.025 ***	0.186 ***	0.024 ***	0.016	0.027 ***
	(0.045)	(0.005)	(0.063)	(0.005)	(0.070)	(0.005)
CONTROL	YES	YES	YES	YES	YES	YES
PROVINCE	YES	YES	YES	YES	YES	YES
YEAR	YES	YES	YES	YES	YES	YES
N	390	390	390	390	390	390
R2	0.058	0.272	0.061	0.290	0.188	0.276
F	2.429	18.183	3.029	21.034	16.387	22.175
p	0.015	0.000	0.003	0.000	0.000	0.000

Note: Robust standard errors are in parentheses, *** and ** indicate significant levels at 1% and 5%, respectively.

. The regression coefficient of II is significantly positive in Model (1), and the regression coefficient of HTA is significantly positive in Model (2). These findings indicate that II can attract the agglomeration of high-tech enterprises, thereby promoting ICIC. This mechanism reveals how II reshapes regional industrial layouts, with important implications for the spatial organization of ICIC. The reason is that the widespread application of industrial robots generates sustained and large-scale demand for intelligent technologies, R&D services, high-end equipment, and system solutions. Such demand is highly attractive to high-tech enterprises engaged in AI, big data, intelligent equipment, and other related fields, effectively driving their regional agglomeration. These enterprises will form close industrial linkages with local manufacturing enterprises and producer services, promote the rapid transformation of innovative achievements and timely feedback of industrial demands, and effectively connect the industrial chain and innovation chain, thereby promoting ICIC. Thus, Hypothesis 2a is verified.

The regression coefficient of II is significantly positive in Model (3), and the regression coefficient of HSA is significantly positive in Model (4). These findings indicate that II can attract the agglomeration of high-skilled labor, thereby promoting ICIC. This mechanism highlights the key role of human capital matching in the integration of industrial and innovation activities, which is crucial for transforming technological advantages into industrial competitiveness. Regions with a higher level of II are characterized by the agglomeration of technology-intensive enterprises. Such enterprises can provide broader career development prospects, more systematic skill-upgrading channels, and more competitive compensation packages for high-skilled labor, thereby driving the agglomeration of high-skilled labor. High-skilled labor can rapidly apply technological achievements from the innovation chain to production, design, operation, and other links in the industrial chain, shortening the transformation cycle of innovative outputs [78]. Meanwhile, the flow of high-skilled labor between the industrial chain and the innovation chain can drive the R&D direction of the innovation chain to align with industrial demand [79], facilitate the precise matching of innovation resources with industrial needs, and further promote ICIC. Thus, Hypothesis 2b is verified.

The regression coefficient of II is positive but not significant in Model (5), whereas the regression coefficient of TI is significantly positive in Model (6). Hypothesis 2c is thus not supported. This insignificant result is economically meaningful: it suggests that the link between II and technological innovation is not immediate in the Chinese context. The possible reasons for this result can be elaborated as follows. First, II in China is still in the stage of “application-oriented popularization” rather than “innovation-driven upgrading”. Most enterprises focus on the procurement of intelligent equipment, the upgrading of existing digital production systems, and the transformation of traditional production processes. These investments are mainly aimed at improving production efficiency, reducing labor costs, and solving practical pain points in the production process, rather than supporting or engaging in technological innovation. As a result, II mainly plays a role in optimizing production links, and it is difficult to directly drive technological innovation, leading to an insignificant direct impact on TI. Second, the link between II and technology innovation needs to be gradually transmitted through channels such as the agglomeration of high-tech enterprises and high-skilled labor [80]. Specifically, the application of industrial robots first attracts high-tech enterprises engaged in intelligent technology R&D and high-skilled talents proficient in intelligent equipment operation and maintenance to gather in the region; then, through technical cooperation, talent exchange, and knowledge spillover between these agglomerated entities and local enterprises, the technological innovation capacity of local enterprises is gradually improved. Such a multi-link transmission process is affected by many factors, such as the absorption capacity of local enterprises, the matching degree of talent structure, and the perfection of regional innovation environment, and thus usually takes a long time to play a role. Therefore, the significant impact of II on technology innovation is difficult to materialize in the short term.

5.2. The Moderating Effect of the Impact of II on ICIC

According to the analysis above, digital infrastructure and marketization influence the relationship between II and ICIC. The regression results are reported in Models (1)–(2) of

Table 8. Moderating effect of the digital infrastructure and marketization.

Variable	(1)	(2)
	ICIC	ICIC
II	0.019 ***	−0.051 **
	(0.006)	(0.022)
DI	0.045
	(0.031)
II × DI	0.024 ***
	(0.009)
MARK		−0.001
		(0.002)
II×MARK		0.008 ***
		(0.002)
CONTROL	YES	YES
PROVINCE	YES	YES
YEAR	YES	YES
N	390	390
R2	0.273	0.306
F	16.945	16.587
p	0.000	0.000

Note: Robust standard errors are in parentheses, *** and ** indicate significant levels at 1% and 5%, respectively.

. The regression coefficient of II × DI is significantly positive in Model (1), and the regression coefficient of II × MARK is significantly positive in Model (2). This indicates that both digital infrastructure and marketization can help strengthen the positive impact of II on ICIC. Thus, Hypotheses 3a,b are verified.

Digital infrastructure serves as the foundation for II. It can effectively reduce information asymmetry and transaction costs in the transformation of innovation achievements, enabling the rapid penetration of intelligent technologies into all links of the industrial chain and innovation chain. This can promote the rapid translation of R&D outcomes into industrial productivity, thereby amplifying the positive effect of II on ICIC. Marketization implies a sound market competition mechanism, which encourages enterprises to better access and integrate innovation resources and accelerate technological upgrading and the transformation of innovation achievements. In addition, in regions with a higher level of marketization, innovation factors such as talent, technology, and capital flow more freely and efficiently, which is conducive to promoting ICIC.

5.3. The Heterogeneity Test of the Impact of II on ICIC

Manufacturing and service enterprises are important entities in the industrial chain and innovation chain. ICIC relies on local manufacturing and service foundations. Among all sample provinces, a province is regarded as having strong manufacturing (service) capacity if the proportion of its secondary (tertiary) industry value-added in regional GDP exceeds the national median.

The regression results of Models (1)–(4) in

Table 9. Heterogeneity test results.

Variable	(1)	(2)	(3)	(4)
	ICIC	ICIC	ICIC	ICIC
II_strong manufacturing	0.019 **
	(0.008)
II_weak manufacturing		0.031 ***
		(0.009)
II_strong service			0.035 ***
			(0.008)
II_weak service				0.020 ***
				(0.008)
CONTROL	YES	YES	YES	YES
PROVINCE	YES	YES	YES	YES
YEAR	YES	YES	YES	YES
N	195	194	192	195
R2	0.322	0.481	0.389	0.334
F	8.371	20.993	11.764	12.540
p	0.000	0.000	0.000	0.000

Note: Robust standard errors are in parentheses, *** and ** indicate significant levels at 1% and 5%, respectively.

show that the regression coefficients of II are significantly positive in both strong and weak manufacturing (service) provinces. This indicates that II can significantly promote ICIC, regardless of each province’s manufacturing (service) foundation. Meanwhile, we find that the regression coefficient of II is larger in weak manufacturing (strong service) provinces. This suggests that the marginal contribution of II is greater in provinces with a weak manufacturing (strong service) foundation. The reason may be that provinces with a weak manufacturing foundation often suffer from shortages of talent and capital, as well as inadequate industrial chain support. II can replace labor and traditional factors with technology, thereby more effectively promoting industrial chain synergy and innovation efficiency. In contrast, provinces with a strong service foundation usually have a well-developed producer services system that provides better supporting services for industrial intelligence. Such supporting services can effectively reduce the cost of II, accelerate technology penetration and the transformation of innovative achievements, and strengthen the promoting effect of II on ICIC.

6. Discussion and Conclusions

II has emerged as a crucial catalyst for fostering high-quality economic growth in China, and also acts as an essential driving force for advancing industrial transformation and achieving the goals of environmental and economic sustainable development. Focusing on the integration of artificial intelligence and production, it provides a new pathway for Chinese manufacturing enterprises to achieve carbon emission reduction targets. The construction of an II-centered industrial value co-creation network is key to enhancing the international competitiveness of industrial chains and consolidating the foundational support for regional sustainable economic development. Accordingly, this study explores the effects and influencing mechanisms of II on ICIC, with its conclusions further offering theoretical support for the practice of sustainable development. The specific findings are presented as follows.

First, II plays a significantly positive role in promoting ICIC, and this conclusion remains valid after a variety of robustness tests. This is because II can break down information barriers between the industrial chain and the innovation chain, shorten the transformation cycle of innovative achievements, and provide a solid industrial foundation and sound ecological support for the integration of the two chains. It also echoes the views of Xie et al. (2024) [38] and Li et al. (2025) [39], who argue that intelligent technologies such as the IoT contribute to ICIC. Different from existing studies that mostly remain at the level of theoretical deduction, this study further complements and enriches the relevant literature by providing rigorous empirical evidence to quantitatively verify the relationship between II and ICIC.

Second, the mediation effect test shows that II can promote ICIC through two paths: the agglomeration of high-tech enterprises and high-skilled labor. This finding aligns with the capital-skill complementarity hypothesis and the theory of external economies. According to the capital-skill complementarity hypothesis, new capital goods induced by II (such as intelligent equipment and software) are highly complementary to high-skilled labor. II increases the demand for intelligence-related producer services and other high-tech enterprises. Meanwhile, it accelerates the upgrading of labor skills, promoting workforce quality advancement. This dual effect not only stimulates stronger market demand for high-skilled talents, but also attracts the continuous inflow and spatial agglomeration of high-tech enterprises and high-skilled labor. Furthermore, from the perspective of the theory of external economies, the agglomeration of high-tech enterprises and high-skilled labor generates labor pooling, shared intermediate inputs, and enhanced knowledge spillovers. These external economies reduce inter-firm cooperation costs, strengthen incentives for R&D investment, and further optimize the linkage and coordination between the industrial chain and the innovation chain. This study opens the black box of how II promotes ICIC. Unlike prior research that neglects the mediating role of high-tech enterprise and high-skilled labor agglomeration, this paper clarifies the underlying influence mechanism more explicitly.

Third, the moderating effect analysis shows that both digital infrastructure and marketization positively moderate the relationship between II and ICIC, further enriching research on the boundary conditions of II in promoting ICIC. As the fundamental carrier of II, digital infrastructure breaks geographical barriers and reduces information asymmetry, enabling intelligent technologies to penetrate all links of industrial and innovation chains more quickly and thereby amplifying the positive impact of II on ICIC. This is consistent with the findings of Wu et al. (2026) [81], who emphasize the supporting role of digital infrastructure in industrial digital transformation. By improving the competition mechanism, optimizing factor allocation efficiency, and strengthening intellectual property protection, marketization provides a sound institutional environment for II to facilitate ICIC. This supports the view of Fan et al. (2019) [65] and Ren et al. (2024) [77] that marketization is conducive to efficient resource allocation and innovative development. These results indicate that the promotional effect of II on ICIC does not operate in isolation but is constrained by the level of regional digital construction and institutional environment, offering important implications for optimizing the practical implementation effect of II.

Fourth, the heterogeneity test shows that the positive impact of II on ICIC is universally applicable; it significantly promotes ICIC in provinces with both strong and weak manufacturing and service industries. This finding is of great practical significance, indicating that II can serve as a universal pathway to advance ICIC regardless of regional industrial foundations. Furthermore, the marginal effect of II is stronger in provinces with weak manufacturing endowments, which can be explained by the late-development advantage. Regions with underdeveloped manufacturing often face shortages of traditional factors such as labor and capital. As II can substitute traditional factors with technology, it more effectively boosts industrial chain synergy and innovation efficiency. By contrast, provinces with strong service industries possess a mature producer service system that provides sound supporting conditions for II, thereby strengthening its promotional effect on ICIC. This finding enriches the research on the regional heterogeneity of ICIC and provides targeted policy implications for different regions to facilitate ICIC via II.

The findings not only respond to the practical demand for breaking external technological blockades and promoting industrial upgrading in China, but also fill the research gap in the quantitative analysis of the relationship between II and ICIC, providing valuable theoretical and practical insights for economies facing similar challenges in industrial transformation and innovation integration.

7. Implications and Limitations

7.1. Implications

First, based on conclusions 1 and 2, the government should attach great importance to the strategic value of II and take it as an important driving force to promote ICIC. The government could rely on the development of II to boost the growth of producer services in the smart technology sector, and thus provide broader ecological support for ICIC.

Second, based on conclusion 2, the government should strengthen the agglomeration of high-tech enterprises and high-skilled labor. On the one hand, the government could increase policy support for high-tech enterprises to accelerate the application of their innovative achievements. On the other hand, the government could improve the training, introduction, and incentive policies for high-skilled labor, promote their rational flow between the industrial chain and innovation chain, and thus achieve the precise matching of innovation resources and industrial demands.

Third, based on conclusion 3, the government should improve the supporting environment to amplify the positive effect of II on ICIC. On the one hand, the government could accelerate the construction of digital infrastructure to amplify its promoting effect on ICIC. On the other hand, the government could continuously promote market-oriented reforms to improve the efficiency of factor allocation and provide a sound institutional environment and market momentum for II to empower ICIC.

Fourth, based on conclusion 4, the government should adopt differentiated development strategies based on regional endowment differences. For provinces with weak manufacturing foundations, they could focus on addressing shortcomings in the intelligent transformation of traditional industries, increase investment in II, so as to rapidly strengthen weak industrial bases. For provinces with strong manufacturing foundations, they could increase R&D investment in core technologies, so as to promote the deep integration of the industrial and innovation chains. For provinces with strong service sectors, they could rely on a sound producer services system to enhance industrial synergy and better release the empowering effect of II. For provinces with weak service sectors, they could accelerate the development of producer services, improve supporting conditions, prioritize R&D, information, and technical services, and provide comprehensive support for the application of II and ICIC.

7.2. Limitations

This study provides valuable insights for understanding the impact of II on ICIC. However, it has several limitations. First, this paper conducts empirical research using provincial data. Future studies could be carried out using enterprise samples to further capture the impact of II on ICIC. Second, this paper identifies the mediating role of high-tech enterprise agglomeration and high-skilled labor agglomeration, but other transmission mechanisms may deserve further exploration. Third, this paper finds that II can significantly promote ICIC in both strong and weak manufacturing (service) provinces. Nevertheless, this study does not consider the influence of institutional environment differences across countries on the empirical results. Future research could conduct further empirical tests using samples from different countries around the world.