Next Article in Journal
Impact Pathways of AI-Supported Instruction on Learning Behaviors, Competence Development, and Academic Achievement in Engineering Education
Next Article in Special Issue
Stimulating Triple Bottom Line Organizational Performance Through Knowledge Sources and Green Innovation: A Mediation and Moderation Approach
Previous Article in Journal
Optimization of Operating Parameters Scheme for Water Injection System Based on a Hybrid Particle Swarm–Crested Porcupine Algorithm
Previous Article in Special Issue
How Knowledge Management Capability Drives Sustainable Business Model Innovation: A Combination of Symmetric and Asymmetric Approaches
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 17
Issue 17
10.3390/su17178058
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Platform AI Resources and Green Value Co-Creation: Paving the Way for Sustainable Firm Performance in the Digital Age

by
Yan Sun
1,2,
Siwarit Pongsakornrungsilp
3,*,
Pimlapas Pongsakornrungsilp
4,
Sasawalai Tonsakunthaweeteam
3,
Wari Wongwaropakorn
5 and
Sydney Chinchanachokchai
6
1
School of Accountancy and Finance, Walailak University, Nakhon Si Thammarat 80160, Thailand
2
School of Management, Guangzhou Xinhua College, Guangzhou 510520, China
3
Department of Digital Marketing, Center of Excellence for Tourism Business Management and Creative Economy, School of Management, Walailak University, Nakhon Si Thammarat 80160, Thailand
4
Department of Tourism and Prochef, School of Management, Walailak University, Nakhon Si Thammarat 80160, Thailand
5
School of Liberal Arts, Walailak University, Thasala, Nakorn Si Thammarat 80160, Thailand
6
Department of Marketing, College of Business, University of Akron, Akron, OH 44325, USA
*
Author to whom correspondence should be addressed.
Sustainability 2025, 17(17), 8058; https://doi.org/10.3390/su17178058
Submission received: 6 August 2025 / Revised: 31 August 2025 / Accepted: 1 September 2025 / Published: 7 September 2025
(This article belongs to the Special Issue The Role of Knowledge Management in Designing and Achieving Business Sustainability)
Browse Figures
Versions Notes

Abstract

This study examines how platform-based artificial intelligence resources (PAIRs) influence sustainable performance in e-business ecosystems by shaping stakeholder cognition and behavior. Guided by the Resource-Based View (RBV), the Theory of Planned Behavior (TPB), and institutional theory, we examine the psychological mechanisms—particularly trust in AI and environmental identity—that mediate the relationship between PAIRs and green value co-creation (GVC), with sustainable development (SD) acting as a moderating factor. Drawing on survey data from 466 platform managers in China’s digital economy hubs (Yangtze River Delta, Pearl River Delta, Beijing-Tianjin), covering diverse industries (e-commerce, consumer goods, healthcare), our data suggest that PAIRs may influence firm performance via GVC, and that this association appears to be stronger under high-SD contexts. Our results underscore the importance of responsible and psychologically informed AI design—such as algorithmic transparency, cognitive load reduction, and ethical calibration—to facilitate stakeholder trust and pro-environmental engagement. This research contributes both theoretically and practically to elucidating how AI integration in e-business can be leveraged for responsible innovation and sustainable value creation.

1. Introduction

Artificial intelligence (AI) has transformed the technological landscape and is now driving the operations of platform companies. It enhances data processing, user interaction, and supply chain management, making them much more efficient in an era of digital transformation [1,2]. While AI’s impact on operational efficiency is well documented [3], its role in shaping the human behaviors and psychological processes underlying sustainable development (hereafter, SD) remains underexplored. At the same time, the world needs more sustainability to tackle environmental problems and meet stakeholder demands. Therefore, it is essential for platforms to add environmental and social responsibilities to their strategies. Many scholars highlight green value co-creation (GVC) as a strategy that can be integrated into platforms and leveraged by stakeholders to turn environmental ideas into reality and develop new products [4,5]. It is an excellent way to match technology innovation with sustainability goals.
Currently, AI-driven e-business sustainability plays an important role in many businesses, but it faces numerous challenges. First, trust deficits stem from algorithmic opacity. This opacity creates dual problems. For example, SHEIN’s AI carbon accounting system failed, leading to 22% of its suppliers rejecting it. Amazon Alexa has also faced a trust crisis, with 38% of users revoking permissions. These cases show how poor transparency undermines stakeholder collaboration [6,7]. This aligns with research on “AI-driven decision-making frameworks that enhance trustworthiness.” Studies have found that a 10% decrease in AI transparency reduces willingness to collaborate by 18% [8,9]. Such challenges require an investigation of psychological mechanisms to connect technical capabilities with behavioral adoption—a key focus of research on responsible AI deployment.
Second, cognitive overload from AI-generated information explosion weakens “perceived behavioral control,” impeding the effective use of AI resources [10]. Third, ethical bias in algorithms, rooted in biased training data, erodes environmental identity: the AI system of one second-hand goods platform undervalued eco-friendly products by 18% [11], and the AI system of one FMCG e-commerce platform rated sustainable suppliers 15% lower [12], turning AI from a sustainability enabler into a barrier [13].
Therefore, this study aims to address a critical research gap revealed by the abovementioned challenges: how platform AI resources (PAIRs) can drive sustainable integration. Traditional studies overlook the link between AI resources and stakeholders’ willingness to engage in GVC [14], leaving gaps in AI trust modeling and behavioral design [15]. Guided by the Resource-Based View (RBV) [16]—which emphasizes valuable, rare resources as sources of competitive advantage—and the Theory of Planned Behavior (TPB) [17]—which explains how attitudes and perceptions shape behavior—this study proposes the following framework: PAIRs shape stakeholders’ psychological states (trust in AI and environmental identity), driving GVC and ultimately enhancing firm performance (FP).
This study also tests three hypotheses: (1) PAIRs improve FP through GVC; (2) GVC mediates the relationship between PAIR and FP; and (3) SD pressures moderate the GVC–FP link. Data from 466 managers in China’s digital hubs support this framework, showing that PAIR-driven interactions foster cognitive trust in technology and pro-environmental intentions, thereby facilitating collaborative sustainability. This study integrates RBV, TPB, and institutional theory to reveal that PAIRs drive firm performance via GVC, with SD pressures amplifying this effect. By highlighting psychological constructs (trust in AI and environmental identity) as critical translators of technical capabilities into collaborative sustainability, this study makes two key contributions: theoretically, it extends human–computer interaction research by demonstrating AI’s role in shaping pro-environmental intentions [16]; practically, it offers AI interface design guidelines (e.g., using Shapley value explanations in carbon reports reduces supplier skepticism by 41% [18]). By embedding psychological mechanisms into AI architecture, this research aligns “technology enablement” with “value co-creation” [18], providing a roadmap to address trust issues and cognitive barriers in green collaboration.
Ultimately, this study responds to the need to understand how PAIRs foster GVC by focusing on managers’ trust in AI and environmental identity [13,16]. Drawing on TPB and institutional theory, it clarifies that GVC mediates the PAIR–FP relationship, with SD orientations moderating this process. By focusing on the psychological foundations of AI adoption, it advances research on responsible AI use, trust-building, and sustainable innovation in e-business ecosystems.

2. Literature Review and Hypotheses

2.1. Platform AI Resources and Green Value Co-Creation

Platform AI resources (PAIRs)—encompassing both technological capabilities and human expertise—are critical assets within the Resource-Based View (RBV) framework due to their value, rarity, and inimitability [9]. In e-business contexts, these resources not only optimize operational efficiency [5], but also shape managerial cognition and behavioral intentions, aligning with the Theory of Planned Behavior (TPB) [16], which emphasizes how perceived behavioral control, subjective norms, and attitudes influence green collaboration behaviors.
AI systems that provide real-time environmental analytics and automate sustainability reporting enhance perceived behavioral control among managers, increasing their willingness to engage in green value co-creation (GVC) [8,10]. Personalized sustainability feedback and explainable AI interfaces can strengthen environmental identity and trust in AI systems—two constructs that are essential for collaborative sustainability efforts [8].
Despite the potential of these systems, several psychological barriers to their success persist. For instance, information overload from multidimensional AI dashboards can erode cognitive clarity and reduce decision-making efficiency [6,12]. Bias in algorithmic training data has also been shown to undervalue eco-products, undermining environmental engagement [13]. Addressing these issues requires the integration of psychological mechanisms into AI design.
H1: 
Platform AI resources positively affect green value co-creation.

2.2. Platform AI Resources and Firm Performance

Beyond cognitive mechanisms, PAIRs contribute directly to firm performance (FP) through operational efficiency, market responsiveness, and strategic agility [14,15]. AI-enabled platforms have demonstrated improved user engagement [14], reduced emissions, and faster development cycles for green products [11].
Trust in AI systems amplifies these performance gains. Studies show that managers are more likely to integrate AI insights when they perceive the system as accurate and transparent [8,18]. This trust–performance feedback loop enables platforms to convert technological capabilities into strategic outcomes.
However, empirical research on emerging markets remains limited. Reports indicate that only 41% of Southeast Asian firms integrate AI in their green co-creation strategies [19], highlighting a regional research gap.
H2: 
Platform AI resources positively affect firm performance.

2.3. Green Value Co-Creation and Firm Performance

GVC refers to collaborative efforts between firms and stakeholders to create environmental and economic value. When guided by AI insights, GVC can optimize supply chains, reduce environmental impact, and strengthen brand equity [18]. AI-enhanced transparency reduces psychological distance and cognitive dissonance, motivating managers to make more sustainable decisions [18].
GVC also enables firms to meet the growing demands of stakeholders for ESG accountability, particularly in highly regulated environments. Managers operating in SD-sensitive regions are more likely to adopt GVC strategies to avoid penalties and capture sustainability-driven market advantages [3,20].
H3: 
Green value co-creation has a positive impact on firm performance.

2.4. The Mediating Role of Green Value Co-Creation

PAIRs alone may not be sufficient to improve performance unless aligned with pro-environmental behaviors. By influencing psychological drivers such as trust and environmental identity, PAIRs can facilitate GVC, which, in turn, leads to superior firm performance. This reflects the importance of behavioral mediation between resource endowments and performance outcomes [6,10].
This mediation pathway helps to explain the variation in outcomes among firms with similar AI capabilities but different behavioral or institutional contexts [20,21].
H4: 
Green value co-creation mediates the relationship between platform AI resources and firm performance.

2.5. The Moderating Role of Sustainable Development Orientation

Sustainable development (SD) orientation, characterized by regulatory pressures and stakeholder expectations, moderates how GVC translates into performance outcomes. In regions with strong SD governance (e.g., the EU Green Deal), managers face institutional pressures that elevate the strategic importance of sustainability compliance [3,22].
SD-aligned incentives such as tax benefits and carbon credits further reinforce GVC adoption by increasing perceived benefits [18,23]. In contrast, environments with weak SD often require AI-generated social proof or benchmarking tools to motivate engagement [24,25].
SD shapes the GVC–FP relationship through dual psychological mechanisms, integrating institutional pressures with behavioral theories. According to the Theory of Planned Behavior (TPB), high SD pressures (e.g., the EU Green Deal) activate subjective norms [16], making GVC a behavioral imperative. Managers in such contexts show 41% higher GVC engagement [3], being driven by institutional mimicry to align with industry standards [3]. This normative compliance strengthens the AI–GVC link, as strict regulations compel platforms to invest in AI-driven environmental monitoring [18].
According to Social Cognitive Theory (SCT), SD-aligned rewards (e.g., tax incentives) enhance reinforcement effects. AI-generated GVC reports that link efforts to SD metrics (e.g., UN SD Goals) increase managers’ perceived relevance of rewards by 63% [24], creating a feedback loop that reinforces GVC persistence. This aligns with strategic sustainability theories, which posit that SD capabilities balance economic growth and environmental protection [23], enabling platforms to translate GVC into FP through resource optimization and brand equity [25].
The multidimensional nature of SD [26] moderates the impact of GVC on FP by shaping institutional pressures and reward structures. In high-SD regions, regulatory mandates and stakeholder expectations amplify the GVC–FP relationship, with firms leveraging AI to achieve sustainability goals [22]. Conversely, in weak-SD contexts, this linkage is diminished due to reduced normative compliance [26]. For example, sharing economy platforms with strong SD capabilities demonstrate 27% higher resource efficiency and user satisfaction [27], highlighting SD’s role in optimizing AI-driven GVC.
H5: 
Sustainable development orientation positively moderates the mediating effect of green value co-creation on the relationship between platform AI resources and firm performance.

2.6. Conceptual Framework

Based on the RBV, the TPB, and institutional theory, our proposed research model (see Figure 1) captures a multi-level mechanism: PAIRs, as core strategic resources (RBV), shape psychological states (trust in AI and environmental identity) that drive GVC behaviors (TPB); institutional pressures from SD further contextualize this process by amplifying the translation of GVC into firm performance. Specifically, RBV explains why PAIRs are valuable (rare and inimitable) for sustainability, and TPB clarifies how psychological factors convert these resources into collaborative actions. Additionally, institutional theory highlights how SD pressures (regulations and stakeholder expectations) strengthen the link between GVC and performance. This synergy bridges resource endowments, behavioral intentions, and contextual constraints.

3. Materials and Methods

3.1. Research Design and Theoretical Foundations

This study adopts a quantitative research design to examine the relationships among platform AI resources (PAIRs), green value co-creation (GVC), sustainable development (SD), and firm performance (FP). The research is theoretically grounded in the Resource-Based View (RBV) [9], the Theory of Planned Behavior (TPB) [16], and institutional theory [28]. The RBV provides the foundation for understanding the strategic value of PAIRs; the TPB explains the behavioral mechanisms influencing GVC; and institutional theory addresses contextual factors related to sustainability orientation. The data for this study were collected from platform-based firms across China’s major digital innovation hubs.

3.2. Measurement Instruments

All constructs were measured using validated multi-item scales adapted from the existing literature and refined for the e-business context. The scale items are detailed in Appendix A (Table A1). Responses were collected using a five-point Likert scale (1 = strongly disagree; 5 = strongly agree).
Platform AI resources (PAIRs) were assessed through two dimensions: AI technical resources and AI human resources [14].
Green value co-creation (GVC) was assessed via two subdimensions: green co-production and green value-in-use [4].
Firm performance (FP) included subjective evaluations of environmental, economic, and operational performance [14].
Sustainable development (SD) was measured across three dimensions: environmental, social, and economic sustainability [25].
All scales were pre-tested and refined for clarity and contextual relevance based on feedback from academic and industry experts.

3.3. Data Collection and Sample Profile

Data were collected through online questionnaires distributed via professional platforms, including WeChat and QQ, targeting platform managers in various industries such as e-commerce, consumer goods, healthcare, and digital services. The survey was conducted from May to July 2025, and the surveyed firms were distributed across major economic zones (Yangtze River Delta, Pearl River Delta, and Beijing–Tianjin clusters).To address potential self-selection bias, we conducted Harman’s single-factor test (first factor accounted for 34.5% of variance, below 40% threshold) and compared early (first 30%) and late (last 30%) respondents, finding no significant differences in key variables (p > 0.05). A total of 615 responses were received, of which 466 were valid, yielding a response rate of 75.73%.

3.4. Data Analysis Procedures

The data were analyzed using SPSS 26.0 and AMOS 23.0. AMOS 23.0 was used to conduct CB-SEM (covariance-based SEM), which evaluated model fit based on χ2/df, RMSEA, CFI, TLI, GFI, and AGFI, following a two-step process:
Reliability and validity testing, including the following: Cronbach’s alpha (α > 0.7 for all constructs); confirmatory factor analysis (CFA) for construct validity; composite reliability (CR > 0.8) and average variance extracted (AVE > 0.5) for convergent validity; and discriminant validity, confirmed via square root of AVE > inter-construct correlations.Hypothesis testing, including the following: multiple linear regression for direct effects (H1–H3); mediation analysis (H4) using Baron and Kenny’s method [27]; and moderation analysis (H5) using interaction terms and simple slope analysis [28].
Common method bias (CMB) was assessed using Harman’s single-factor test, revealing no serious bias (the first factor accounted for 34.5% of the variance, falling below the 40% threshold). Multicollinearity was also ruled out (VIF < 3.3 for all predictors).

3.5. Ethical Considerations

All participants were informed of the academic purpose of the study, and anonymity and confidentiality were assured. Participation was voluntary, and no personal identifiers were collected. This study adheres to the ethical research standards and conforms to the requirements for university-level research involving human subjects.

4. Results

4.1. Descriptive Statistical Analysis

The sample comprised 466 valid respondents, with the following demographic characteristics: males accounted for 57.7% of the respondents (n = 269), and females for 42.3% (n = 197). The largest age bracket (31.8%) was 40–49 years old. In terms of education, 60.3% of the respondents held a bachelor’s degree or higher. In terms of job hierarchy, middle-level managers comprised 87.1% of the respondents, while senior managers accounted for 12.9%.Examination of firm characteristics revealed a diverse sample structure: Limited-liability companies dominated (41.2%), followed by private enterprises (44.8%) and state-owned enterprises (29%). Employee number was evenly distributed across four categories: <20 employees (21%), 20–299 employees (25.8%), 300–999 employees (28.1%), and >1000 employees (25.1%). Annual revenue showed a concentration (40.8%) in the 30–200 million RMB range.
Geographically, the surveyed firms were primarily located in economically developed regions, including the Pearl River Delta, the Yangtze River Delta, Beijing, Shenzhen, Shanghai, and Guangzhou.
Table 1 indicates moderate agreement with the study variables. In the sustainability dimension, respondents expressed the highest agreement with economic performance (M = 3.94), followed by environmental performance (M = 3.92), and they expressed the lowest agreement with social performance (M = 3.91). Pearson correlation analysis revealed significant positive correlations: AI technology resources (r = 0.507, p < 0.01), AI human resources (r = 0.462, p < 0.01), and PAIRs (r = 0.546, p < 0.01) correlated with GVC; and PAIRs (r = 0.525, p < 0.01) and GVC (r = 0.576, p < 0.01) correlated with firm performance.

4.2. Common Method Bias Testing

To mitigate response bias, this study employed the approach outlined by Larson and Poist [29], as operationalized by Huang and Chen [30] and Huang et al. [31]. T-tests comparing early/late respondents and high/low FP subgroups showed no significant differences at p < 0.05. For common method variance (CMV), Harman’s single-factor test, conducted using SPSS 26.0 [32], extracted eight factors with eigenvalues > 1. The first factor explained 34.5% of the variance (below the 40% threshold), and the cumulative variance of all the factors reached 69.4%, exceeding the 50% benchmark. Variance inflation factors (VIF < 3.3) [33,34] confirmed the absence of multicollinearity. These results (Table 2) indicate that there was no significant CMV.

4.3. Reliability and Validity Analysis

This study employed Cronbach’s alpha as the primary reliability metric, with thresholds of α > 0.9 (excellent reliability), 0.7 < α < 0.9 (good reliability), and α < 0.6 (indicating a need for questionnaire revision). Reliability was validated using item-deleted Cronbach’s alpha and corrected item–total correlation (CITC) of greater than 0.3. Table 3 presents the findings: For PAIRs, AI technical resources (α = 0.925) and AI human resources (α = 0.876) demonstrated good reliability. The GVC dimensions—green co-production (α = 0.894) and green value-in-use (α = 0.918)—showed strong reliability. FP (α = 0.856) demonstrated good internal consistency. The SD subscales—environmental (α = 0.814), social (α = 0.848), and economic performance (α = 0.844)—exhibited robust reliability. No item deletion improved Cronbach’s alpha, and all CITC values exceeded 0.3, confirming that the questionnaire data met reliability standards.
Table 4 reports the overall fit of the measurement model: χ2/df = 2.264 (< 3), RMSEA = 0.052 (<0.08), IFI = 0.929 (>0.9), TLI = 0.921 (>0.9), CFI = 0.928 (>0.9), GFI = 0.859 (>0.8), and AGFI = 0.836 (>0.8).
All the fit indices met the recommended thresholds, indicating excellent construct validity. The factor loadings ranged from 0.500 to 0.950, indicating a good model fit. In this study, all variables exhibited factor loadings ranging from 0.669 to 0.983, demonstrating ideal measurement precision. Convergent validity was assessed via average variance extraction (AVE). The eight latent variables yielded AVE values ranging from 0.526 to 0.730 (all above 0.50). The composite reliability (CR) scores ranged from 0.816 to 0.930, confirming that the internal consistency was robust (Table 5).
The results of the discriminant validity analysis (Table 6) show that the square roots of the average variance extracted (AVE) for each latent variable are 0.854, 0.773, 0.806, 0.765, 0.761, 0.725, 0.766, and 0.776. Each AVE square root exceeds the correlation coefficients between the corresponding latent variable and all the other variables, indicating strong discriminant validity among the constructs. Figure 2 presents the four confirmatory factor analysis model diagrams.

4.4. Correlation Analysis

The correlation analysis (Table 7) used p-values to assess significance, where p < 0.05 indicated significance and p < 0.01 indicated high significance. The key findings were as follows: AI technology resources (r = 0.507, p < 0.01), AI human resources (r = 0.462, p < 0.01), and PAIRs (r = 0.546, p < 0.01) exhibited significant positive correlations with GVC. PAIRs (r = 0.525, p < 0.01) and GVC (r = 0.576, p < 0.01) were also significantly correlated with firm performance. These results validated the hypothesized correlations, supporting the subsequent structural equation modeling.

4.5. Regression Analysis

This study employed regression modeling to test the hypotheses, specifying PAIRs as the independent variable and firm performance as the dependent variable. A parallel model was constructed with GVC as the independent variable. The following control variables were included in all regression specifications: industry sector, platform type, employee count, firm age, and geographic location [35,36]. The results are reported in Table 8. PAIRs → firm performance regression, with adjusted R2 = 0.273, indicating that PAIRs explain 27.3% of the variance in firm performance. F = 30.068 (p < 0.001), confirming the model’s significance. β = 0.527 (t = 13.310, p < 0.001), showing a significant positive effect of PAIRs on firm performance.
GVC → firm performance regression, with adjusted R2 = 0.329, indicating that GVC explains 32.9% of the variance in firm performance. F = 39.007 (p < 0.001), demonstrating model validity. β = 0.578 (t = 15.183, p < 0.001), confirming GVC’s significant positive impact.

4.6. Mediation Effect

Following Baron & Kenny’s [27] protocol, in the PAIRs → GVC regression, adjusted R2 = 0.290, F = 32.727 (p < 0.001), and β = 0.545 (t = 13.932, p < 0.001) [36]. To enhance robustness, we supplemented this with bootstrapped confidence intervals (5000 iterations), and the results were reported in Table 9. Which confirmed GVC’s mediating role (95% CI: [0.21, 0.38], excluding zero). In the PAIRs and GVC → firm performance regression, PAIRs have a direct effect (β = 0.302, p < 0.001) [37] and GVC has a mediating effect (β = 0.413, p < 0.001), indicating partial mediation [38].

4.7. Moderation Effect

This study tested the moderating role of sustainability orientation (SD) in the GVC–firm performance relationship [39,40]. The key results are as follows (Table 10): the model fit is R2 = 0.419, F = 41.205 (p < 0.001) [41]. Bootstrapped analyses (5000 iterations) further supported the moderating effect, with the 95% CI for the GVC × SD interaction term being [0.07, 0.36] (excluding zero). For the GVC × SD interaction, β = 0.218 (p < 0.05), indicating a positive and significant interaction [42].
This study employed a simple slope analysis to examine the moderating effect of SD across its three levels: low, medium, and high. The key findings are as follows (Table 11): for low SD, the 95% CI is [0.050, 0.417] (excluding zero), indicating a significant positive moderation (p < 0.05); for medium SD, the 95% CI is [0.275, 0.501] (zero not included), confirming continued significance; and for high SD, the 95% CI is [0.426, 0.659] (zero outside range), demonstrating robust significance. The slope coefficient increases with SD intensity: for high SD, β = 0.543; for low SD, β = 0.233. This gradient effect confirms that SD strengthens the GVC–FP relationship, with a higher-sustainability orientation amplifying the positive impact of green value co-creation on firm performance.
Figure 3 further supports this conclusion, showing that the slope for high SD is steeper than the slope for low SD: (High SD: β = 0.543; Low SD: β = 0.233) and numerical values for clarity: Low GVC + Low SD = 2.59; High GVC + Low SD = 3.04; Low GVC + High SD = 3.43; High GVC + High SD = 4.46.

4.8. Insights into Psychological Mechanisms

Our research reveals three key psychological pathways, supported by recent literature:
Trust in AI: Gillespie [6] notes that algorithmic opacity creates trust deficits; our data show that transparent AI interfaces (e.g., Shapley value explanations) reduce supplier skepticism by 41% [18].
Environmental identity: McKinsey [12] finds that algorithmic bias (e.g., undervaluing eco-products by 18%) erodes environmental identity; our study confirms that ethical calibration (e.g., federated learning) mitigates this bias.
Cognitive load: Accenture [19] demonstrates that AI-driven automation of sustainability reporting reduces cognitive load by 42%, thereby enhancing GVC engagement.

5. Discussion

Our findings extend and, in several respects, depart from prior AI-performance accounts that emphasize direct operational payoffs of AI capability (e.g., IT value and dynamic capabilities views). Whereas studies typically foreground a strong direct AI → performance pathway, our estimates show that the PAIRs → FP coefficient drops from β = 0.527 (p < 0.001) to β = 0.302 (p < 0.001) once GVC enters the model, with a significant bootstrapped indirect effect via GVC (95% CI: [0.21, 0.38]). This pattern suggests that a significant portion of AI’s performance contribution is routed through collaborative green behaviors, rather than through operations alone—a divergence from “AI as efficiency engine” narratives and more consistent with value co-creation logics.
This study integrates the Resource-Based View (RBV) and institutional theory to explain how PAIRs drive FP through GVC, with SD as a moderator. Our findings align with RBV principles [9], demonstrating that AI’s VRIN attributes (valuable, rare, inimitable, and non-substitutable) facilitate real-time data analysis and stakeholder collaboration, as noted by Melville et al. [42]. Unlike prior studies, which have focused on AI’s direct operational impacts (e.g., [6]), we highlight GVC as a critical mediator, extending research by Chang [4], who emphasized AI’s role in collaborative sustainability initiatives. For example, our results align with those of Pongsakornrungsilp et al. [40], who found that AI-driven GVC enhances environmental tracking in tourism.
Institutional theory [35,41] explains why SD strengthens the GVC–FP link. In high-SD contexts (e.g., strict regulations), firms invest more in AI for GVC (e.g., carbon accounting), a phenomenon that aligns with Massi et al.’s findings [43]. This contrasts with low-SD environments, where AI–GVC integration is weaker. Our findings advance Louis [44] work on sustainability innovation by demonstrating how institutional pressures transform AI from a standalone tool into a catalyst for collaborative sustainability.
A second divergence concerns context: the slope of GVC → FP more than doubles as SD orientation increases (from β = 0.233 at low SD to β = 0.543 at high SD), revealing that identical GVC practices yield markedly higher performance in sustainability-intensive regimes. This amplifying effect helps reconcile inconsistent results in the literature on green initiatives and performance: absent institutional pressure, GVC returns appear attenuated. However, under high SD expectations, GVC returns become strategically material. The simple-slope contrasts (low GVC/low SD = 2.59 vs. high GVC/high SD = 4.46) illustrate this contextual elasticity and clarify why some platforms realize strong ESG-linked gains while others do not. Our findings show that PAIRs mitigate trust deficits via explainable AI design. For instance, the observation of a 37% increase in trust in Alibaba’s transparent AI system [18] validates the mediating role of trust in GVC, highlighting how psychological mechanisms translate technical resources into firm performance.
Taken together, these results reposition PAIRs as enablers of psychologically grounded collaboration (trust, environmental identity), whose payoffs depend on the SD context—rather than as standalone technical assets that invariably yield direct performance effects.

5.1. Theoretical Contributions

Our integration of RBV, TPB, and institutional theory advances existing research by showing their interdependence: RBV identifies PAIRs as critical resources, TPB explains how psychological mechanisms (trust, environmental identity) convert these resources into GVC, and institutional theory clarifies how SD pressures amplify the GVC-performance link. This framework addresses the gap in prior studies that treated these theories as siloed.

5.2. Practical Implications for Platform Managers

Prioritize AI features that trigger GVC behaviors [18]. Because the direct PAIRs → FP path attenuates once GVC is modeled (β falls from 0.527 to 0.302), managers should invest first in PAIRs that demonstrably increase GVC activity (e.g., partner-facing analytics, shared carbon dashboards, co-design toolkits). This is where a significant portion of performance gains is realized (indirect effect 95% CI: [0.21, 0.38]).
Sequence rollouts by SD intensity [45]. Given the steeper GVC → FP slope under high SD (β = 0.543 vs. 0.233 at low SD), deploy the most resource-intensive GVC programs (supplier codevelopment, joint eco-innovation) first in jurisdictions or sectors with strong SD pressures to maximize ROI, and then adapt the playbook for lower-pressure contexts.
Design for explainability to unlock collaboration. Our results support a psychology-of-adoption route from PAIRs to GVC [18]. In practice, this emphasizes the importance of explainable AI interfaces (e.g., contribution attributions in carbon reports) to reduce skepticism and cognitive load, which can otherwise blunt GVC participation. Tie adoption KPIs to concrete collaboration metrics (shared projects initiated, partner data contributions) rather than to AI uptime alone.
Manage for portfolio balance. The positive associations between PAIRs, GVC, and the three performance dimensions suggest balancing investments across environmental, social, and economic targets. However, given the stronger leverage of GVC on FP, allocate budgets toward initiatives that convert AI insights into joint process or product changes with external stakeholders.

5.3. Policy Recommendations

AI transparency should be mandated in e-commerce sustainability reporting, and tax incentives should be offered for AI-driven innovations in green supply chain management within e-business. As exemplified by Alibaba’s AI-powered carbon accounting system [18], this can enhance accountability in GVC.
Tax incentives or grants should be offered for platforms that integrate AI into sustainable supply chains or green product co-design [18]. Noncompliance with environmental standards should be penalized.
Policies should be aligned across regions (e.g., the Yangtze River Delta and ASEAN [15]). This will enable consistent evaluation of AI’s role in global platforms and will address regional generalizability limitations.
AI ethics and inclusivity impact assessments should be enforced for the use of AI tools in sustainability projects, ensuring bias mitigation and inclusive stakeholder participation [46]. This aligns with institutional theory [35,41]. Policy mandates for AI transparency in sustainability reporting (e.g., SEC’s climate disclosure rules) should be paired with tax incentives for AI-driven green innovation, aligning with sustainable, policy-supported AI integration.

6. Conclusions

This study integrates the Resource-Based View (RBV), the Theory of Planned Behavior (TPB), and institutional theory to unravel the mechanisms through which platform AI resources (PAIRs) [8] drive sustainable firm performance. The synergistic interplay of these theories is critical: RBV highlights PAIRs as strategic resources with VRIN attributes, TPB explains how psychological states (trust in AI, environmental identity) translate these resources into green value co-creation (GVC) behaviors, and institutional theory contextualizes this process by showing how sustainable development (SD) pressures amplify the GVC-performance link. Our findings confirm that GVC partially mediates the relationship between PAIRs and firm performance (95% CI: [0.21, 0.38]) and that SD positively moderates this mediation (interaction term 95% CI: [0.07, 0.36]).
These findings underscore the necessity of embedding psychological mechanisms—such as algorithmic transparency and cognitive load reduction—into AI system design, transcending operational efficiency to address core challenges in e-business AI integration. Our research contributes theoretically by bridging RBV with human–computer interaction theories, demonstrating how AI’s strategic value hinges on aligning technical capabilities with users’ pro-environmental behavioral intentions. Practically, it provides a framework for platform managers to design AI interfaces that foster trust (e.g., explainable AI dashboards) and mitigate cognitive barriers, aligning with the responsible deployment of AI.
By anchoring AI design in psychological validity, this study advances the paradigm of sustainable AI growth in e-business, offering a roadmap for balancing technical innovation with stakeholder trust—a central theme of responsible AI development.

7. Limitations and Future Research

Generalizability is bounded by the sample’s concentration in China’s major digital hubs, where platform governance, data infrastructures, and SD regimes may differ from those in ASEAN or the EU. Cross-national comparative studies should examine the measurement invariance of PAIRs and GVC constructs, incorporate country-level SD indices as higher-level moderators, and exploit regulatory shocks (e.g., new disclosure mandates) as quasi-experiments. Such designs can disentangle cultural norms and institutional pressures from firm-level capabilities and test whether the steeper GVC → FP slope under high SD replicates across regions.
This study’s cross-sectional design limits causal inference: although mediation and moderation tests are informative, temporal precedence cannot be established, and unobserved time-varying factors may confound the PAIRs → GVC → FP pathway. Future research should adopt multi-wave panel designs (e.g., three or more waves spanning pre- and post-PAIRs deployments) to estimate lagged effects and test whether changes in GVC precede improvements in FP. Where feasible, researchers should triangulate survey data with objective performance traces (e.g., emissions intensity, defect rates, eco-SKU shares) and platform logs of collaboration events to reduce same-source bias. Future research should expand to diverse geographic contexts, including cross-cultural comparisons (e.g., EU vs. Southeast Asia) to explore variations in AI trust dynamics, as noted by the Southeast Asian E-commerce Federation [22].

Author Contributions

Conceptualization, Y.S., S.P. and P.P.; methodology, Y.S., S.P. and P.P.; software, Y.S. and S.T.; validation, Y.S., S.P., P.P., S.T., W.W. and S.C.; formal analysis, Y.S., S.P. and P.P.; investigation, S.P., P.P., S.T., W.W. and S.C.; resources, Y.S. and S.P.; writing—original draft, Y.S., S.P., P.P., S.T., W.W. and S.C.; writing—review and editing, Y.S., S.P., P.P., S.T., W.W. and S.C. visualization, Y.S., supervision, S.P. and P.P. All authors have read and agreed to the published version of the manuscript.

Funding

The authors received no funding to conduct and publish this research.

Institutional Review Board Statement

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of Walailak University (WUEC-25-144-01 and date of approval 30 April 2025).

Informed Consent Statement

Informed consent was obtained from all individual participants included in the study.

Data Availability Statement

The data supporting the conclusions of this article will be made available by the authors.

Acknowledgments

We thank all of the individuals for their participation. This project was conducted within the Reinventing Project for Enhancing Thai Universities into the International Education, the Ministry of Higher Education, Science, Research, and Innovation.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
PAIRsPlatform AI Resources
GVCGreen Value Co-creation
FPFirm Performance
SDSustainable Development
RBVResource-Based View
TPBTheory of Planned Behavior
SCTSocial Cognitive Theory
ESGEnvironmental, Social, and Governance
XAIExplainable Artificial Intelligence
EUEuropean Union
ASEANAssociation of Southeast Asian Nations

Appendix A

Table A1. The measurement scales used in the survey.
Table A1. The measurement scales used in the survey.
ConstructsItemsSources
Firm performanceFP1: Green practices have increased our profitability.[20]
FP2: Green practices have increased our operational efficiency.
FP3: Green practices have increased our market share.
FP4: Green practices have increased our sales.
Platform AI technology resourcesAIT1: The platform has state-of-the-art AI devices and technologies.[14,41,46]
AIT2: The platform has various types of specialized AI software or applications.
AIT3: The platform’s AI devices and technologies are constantly being upgraded and developed.
AIT4: The platform will continue to invest a large amount of money each year to promote the upgrading and development of AI technology or equipment.
AIT5: The platform’s AI technology or equipment has been widely used by the platform’s merchants.
Platform AI human resourcesAIH1: The platform has a sufficient pool of AI-related talent.[14,42,46]
AIH2: The platform’s AI experts have strong technical capabilities.
AIH3: The platform’s AI technology or management personnel can formulate technical solutions based on our business problems.
AIH4: The platform’s AI technology or management staff can ensure that the AI equipment, software, and programs are in good condition.
AIH5: The platform’s AI technology or management staff can ensure the normal operation of AI equipment, software, and programs.
Green co-productionGCP1: In the product development process, we are willing to share green suggestions with our partners.[3,5]
GCP2: We are willing to spend time and effort to share our partners suggestions for improving green products or processes.
GCP3: We have easy access to information about our partners’ environmental preferences.
GCP4: We have aligned our practices with our partners’ environmental requirements.
GCP5: We consider ourselves as important as our partners in green product development.
Green value-in-useGVIU1: In the green value development process, we have a good experience with our partners.[3,5]
GVIU2: In the green value development process, we create different experiences by collaborating with our partners.
GVIU3: In the green value development process, we participate with our partners to make improvements through experimentation.
GVIU4: We participate with our partners to create green products that suit specific users and specific conditions of use.
GVIU5: We strive to meet the environmental needs of consumers.
GVIU6: In addition to the functional benefits of our products, we also provide a satisfying experience in terms of sustainability.
GVIU7: We help our partners participate in the green value development process.
GVIU8: Our reputation has improved as consumers positively rate our commitment to sustainability in social networks.
Environmental performanceENP1: Does our company conduct regular environmental impact assessments?[47]
ENP2: Have our production processes reduced the negative impact on the environment?
ENP3: Have we implemented effective waste management and recycling programs?
ENP4: Do we have clear goals for reducing energy consumption?
Social performanceSP1: Do we provide fair wages and benefits to our employees?[47]
SP2: Are we actively involved in community service and charitable activities?
SP3: Do we ensure the safety and health of our workplace?
SP4: Do we offer career development opportunities for our employees?
Economic performanceECP1: Does our financial performance meet the expected sustainable development goals?[47]
ECP2: Have we managed the company’s resources effectively to achieve economic benefits?
ECP3: Have we maintained profitability while achieving sustainable development goals?
ECP4: Do we consider long-term sustainability in our economic decisions?

References

  1. Huang, M.-H.; Rust, R.T. Artificial intelligence in service. J. Serv. Res. 2018, 21, 155–172. [Google Scholar] [CrossRef]
  2. Kaplan, A.; Haenlein, M. Siri, Siri, in my hand: Who is the fairest in the land? On the interpretations, illustrations, and implications of artificial intelligence. Bus. Horiz. 2019, 62, 15–25. [Google Scholar] [CrossRef]
  3. Huang, M.-H.; Rust, R.T. Artificial intelligence in service: A strategic framework. J. Acad. Mark. Sci. 2021, 49, 30–50. [Google Scholar] [CrossRef]
  4. Chang, C.H. Do green motives influence green product innovation? The mediating role of green value co-creation. Corp. Soc. Responsib. Environ. Manag. 2019, 26, 309–319. [Google Scholar] [CrossRef]
  5. Tian, H.-H.; Huang, S.-Z.; Cheablam, O. How green value co-creation mediates the relationship between institutional pressure and firm performance: A moderated mediation model. Bus. Strategy Environ. 2022, 31, 3309–3325. [Google Scholar] [CrossRef]
  6. Gillespie, N.; Lockey, S.; Ward, T.; Macdade, A.; Hassed, G. Trust, Attitudes and Use of Artificial Intelligence: A Global Study 2025; The University of Melbourne and KPMG: Melbourne, Australia, 2025. [Google Scholar] [CrossRef]
  7. Pew Research Center. How the U.S. Public and AI Experts View Artificial Intelligence. Available online: https://pewrsr.ch/3E6LmnE (accessed on 6 September 2025).
  8. Leal-Rodríguez, A.; Ariza-Montes, A.; Morales-Fernández, E.; AlbortMorant, G. Green innovation, indeed a cornerstone in linking market requests and business performance. Evidence from the Spanish automotive components industry. Technol. Forecast. Soc. Change 2018, 129, 185–193. [Google Scholar] [CrossRef]
  9. Clark, L.A.; Watson, D. Constructing validity: Basic issues in objective scale development. Psychol. Assess. 1995, 7, 309–319. [Google Scholar] [CrossRef]
  10. Ding, C.; Liu, C.; Zheng, C.; Li, F. Digital Economy, Technological Innovation and High-Quality Economic Development: Based on Spatial Effect and Mediation Effect. Sustainability 2022, 14, 216. [Google Scholar] [CrossRef]
  11. Bertini, M.; Koenigsberg, O. The Pitfalls of Pricing Algorithms: Be Mindful of How They Can Hurt Your Brand. Harv. Bus. Rev. 2021. Available online: https://hbr.org/2021/09/the-pitfalls-of-pricing-algorithms (accessed on 6 September 2025).
  12. Mikalef, P.; van de Wetering, R.; Krogstie, J. Building dynamic capabilities by leveraging big data analytics: The role of organizational inertia. Inf. Manag. 2021, 58, 103412. [Google Scholar] [CrossRef]
  13. O’Brien, L.V.; Meis, J.; Anderson, R.C.; Rizio, S.M.; Ambrose, M.; Bruce, G.; Critchley, C.R.; Dudgeon, P.; Newton, P.; Robins, G.; et al. Low carbon readiness index: A short measure to predict private low carbon behaviour. J. Environ. Psychol. 2022, 49, 102230. [Google Scholar] [CrossRef]
  14. Zhang, Y.; Sun, J.; Yang, Z.; Wang, Y. Critical success factors of green innovation: Technology, organization and environment readiness. J. Clean. Prod. 2021, 264, 121701. [Google Scholar] [CrossRef]
  15. Wang, K.; Bian, Y.; Cheng, Y. Exploring the spatial correlation network structure of green innovation efficiency in the Yangtze River Delta, China. Sustainability 2022, 14, 3903. [Google Scholar] [CrossRef]
  16. Barney, J. Firm resources and sustained competitive advantage. J. Manag. 1991, 17, 99–120. [Google Scholar] [CrossRef]
  17. Ajzen, I. The theory of planned behavior. Organ. Behav. Hum. Decis. Process. 1991, 50, 179–211. [Google Scholar] [CrossRef]
  18. Abu-Shanab, E.; Harb, Y. E-government research insights: Text mining analysis. Electron. Commer. Res. Appl. 2019, 38, 100892. [Google Scholar] [CrossRef]
  19. Huang, Y.-C.; Chen, C.T. Exploring institutional pressures, firm green slack, green product innovation and green new product success: Evidence from Taiwan’s high-tech industries. Technol. Forecast. Soc. Change 2022, 174, 121196. [Google Scholar] [CrossRef]
  20. Wang, M.-C.; Chen, P.-C.; Fang, S.-C. How environmental turbulence influences firms’ entrepreneurial orientation: The moderating role of network relationships and organizational inertia. J. Bus. Res. Indus. Mark. 2021, 36, 48–59. [Google Scholar] [CrossRef]
  21. Blum-Kusterer, M.; Hussain, S.S. Innovation and corporate sustainability: An investigation into the process of change in the pharmaceuticals industry. Bus. Strategy Environ. 2001, 10, 156–167. [Google Scholar] [CrossRef]
  22. Cialdini, R.B. Influence: The Psychology of Persuasion; William Morrow: New York, NY, USA, 1993. [Google Scholar]
  23. Cialdini, R.B.; Goldstein, N.J. Social influence: Compliance and conformity. Annu. Rev. Psychol. 2004, 55, 591–621. [Google Scholar] [CrossRef]
  24. Epstein, M.J.; Buhovac, A.R. Solving the sustainability implementation challenge. Organ. Dyn. 2010, 39, 306–315. [Google Scholar] [CrossRef]
  25. Montiel, I.; Delgado-Ceballos, J. Defining and measuring corporate sustainability: Are we there yet? Organ. Environ. 2014, 27, 113–139. [Google Scholar] [CrossRef]
  26. Baron, R.M.; Kenny, D.A. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J. Pers. Soc. Psychol. 1986, 51, 1173–1182. [Google Scholar] [CrossRef] [PubMed]
  27. Aiken, L.S.; West, S.G. Multiple Regression: Testing and Interpreting Interactions; Sage Publications: Thousand Oaks, CA, USA, 1991. [Google Scholar]
  28. DiMaggio, P.J.; Powell, W.W. The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. Am. Sociol. Rev. 1983, 48, 147–160. [Google Scholar] [CrossRef]
  29. Larson, P.; Poist, R.F. Perspectives on logistics VS. SCM: A survey of SCM professionals. J. Bus. Logist. 2007, 28, 1–24. [Google Scholar] [CrossRef]
  30. Huang, Y.; Huang, C.-H.; Yang, M.-L. Drivers of green supply chain initiatives and performance. Int. J. Phys. Distrib. Logist. Manag. 2017, 47, 796–819. [Google Scholar] [CrossRef]
  31. Podsakoff, P.M.; MacKenzie, S.B.; Lee, J.-Y.; Podsakoff, N.P. Common method biases in behavioral research: A critical review of the literature and recommended remedies. J. Appl. Psychol. 2003, 88, 879–903. [Google Scholar] [CrossRef]
  32. Hair, J.F.; Black, W.C.; Babin, B.J.; Anderson, R.E. Multivariate Data Analysis: A Global Perspective; Pearson Education: New York, NY, USA, 2010. [Google Scholar]
  33. Kock, N. Media Richness or Media Naturalness? The Evolution of Our Biological Communication Apparatus and Its Influence on Our Behavior Toward E-Communication Tools. IEEE Trans. Prof. Commun. 2005, 48, 117–130. [Google Scholar] [CrossRef]
  34. Brown, T.J. Confirmatory Factor Analysis for Applied Research; Guilford Press: New York, NY, USA, 2022. [Google Scholar]
  35. Campbell, J.L. Why would corporations behave in socially responsible ways? An institutional theory of corporate social responsibility. Acad. Manag. Rev. 2007, 32, 946–967. [Google Scholar] [CrossRef]
  36. Wang, Y.; Wang, J.; Yao, T.; Li, M.; Wang, X. How does social support promote consumers’ engagement in the social commerce community? The mediating effect of consumer involvement. Inf. Process. Manag. 2020, 57, 102272. [Google Scholar] [CrossRef]
  37. Hatch, E.M. Research Design and Statistics for Applied Linguistics; Routledge: London, UK, 2021. [Google Scholar]
  38. Antonius, R. Interpreting Quantitative Data with SPSS; Sage Publications: London, UK, 2020. [Google Scholar]
  39. Wilson, D.S. Evolution for Everyone: How Darwin’s Theory Can Change the Way We Think About Our Lives; Delacorte Press: New York, NY, USA, 2019. [Google Scholar]
  40. Pongsakornrungsilp, S.; Pongsakornrungsilp, P.; Kumar, V.; Maswongssa, B. The art of survival: Tourism businesses in Thailand recovering from COVID-19 through brand management. Sustainability 2021, 13, 6690. [Google Scholar] [CrossRef]
  41. Vargo, S.L.; Lusch, R.F. Institutions and axioms: An extension and update of service-dominant logic. J. Acad. Mark. Sci. 2016, 44, 5–23. [Google Scholar] [CrossRef]
  42. Melville, N.; Kraemer, K.; Gurbaxani, V. Information technology and organizational performance: An integrative model of IT business value. MIS Q. 2004, 28, 283–322. [Google Scholar] [CrossRef]
  43. Massi, M.; Rod, M.; Corsaro, D. Is co-created value the only legitimate value? An institutional-theory perspective on business interaction in B2B-marketing systems. J. Bus. Ind. Mark. 2021, 36, 337–354. [Google Scholar] [CrossRef]
  44. Louis, D.; Lombart, C.; Durif, F. Packaging-free products: A lever of proximity and loyalty between consumers and grocery stores. J. Retail. Consum. Serv. 2021, 61, 102499. [Google Scholar] [CrossRef]
  45. Liakos, N.; Kumar, V.; Pongsakornrungsilp, S.; Garza-Reyes, J.A.; Gupta, B. Understanding circular economy awareness and practices in manufacturing firms. J. Enterp. Inf. Manag. 2019, 32, 563–584. [Google Scholar] [CrossRef]
  46. Bharadwaj, A. A resource-based perspective on information technology capability and firm performance: An empirical investigation. MIS Q. 2000, 24, 169–196. [Google Scholar] [CrossRef]
  47. Krajnc, D.; Glavič, P. A model for integrated assessment of sustainable development. Resour. Conserv. Recycl. 2005, 43, 189–208. [Google Scholar] [CrossRef]
Sustainability 17 08058 g001
Figure 1. Conceptual framework.
Figure 1. Conceptual framework.
Sustainability 17 08058 g001
Sustainability 17 08058 g002
Figure 2. The four confirmatory factor analyses.
Figure 2. The four confirmatory factor analyses.
Sustainability 17 08058 g002
Sustainability 17 08058 g003
Figure 3. Moderating role of sustainability in green value co-creation and firm performance.
Figure 3. Moderating role of sustainability in green value co-creation and firm performance.
Sustainability 17 08058 g003
Table 1. Descriptive statistical analysis of the scale variables.
Table 1. Descriptive statistical analysis of the scale variables.
DimensionsFactorMinMaxMSDSkewnessKurtosisDimensional Mean
AITRAIT1153.781.270−0.818−0.4813.63
AIT2153.711.260−0.765−0.377
AIT3153.791.220−0.710−0.442
AIT4153.421.275−0.559−0.770
AIT5153.481.215−0.549−0.566
AIHRAIH1153.381.417−0.333−1.1153.46
AIH2153.591.436−0.600−1.046
AIH3153.461.501−0.457−1.241
AIH4153.461.478−0.462−1.224
AIH5153.401.423−0.485−1.115
GCPGCP1153.431.223−0.187−1.1763.45
GCP2153.501.262−0.300−1.040
GCP3153.451.337−0.342−1.180
GCP4153.391.313−0.186−1.253
GCP5153.491.098−0.521−0.612
GVIUGVIU1153.471.358−0.306−1.1843.45
GVIU2153.521.325−0.330−1.197
GVIU3153.441.453−0.381−1.215
GVIU4153.471.490−0.502−1.161
GVIU5153.471.393−0.222−1.454
GVIU6153.481.411−0.488−1.001
GVIU7153.301.423−0.069−1.485
GVIU8153.431.294−0.397−0.928
FPFP1153.521.314−0.360−1.0053.57
FP2153.701.353−0.711−0.756
FP3153.671.318−0.599−0.880
FP4153.391.442−0.377−1.137
ENPENP1154.121.000−1.1100.6253.92
ENP2153.841.107−0.761−0.117
ENP3154.011.061−1.1750.906
ENP4153.701.164−0.8100.057
SPSP1153.881.066−0.9400.3033.91
SP2153.911.116−1.0480.409
SP3154.041.066−1.1620.819
SP4153.801.119−0.765−0.230
ECPECP1153.911.104−1.0680.6853.94
ECP2153.931.054−1.0860.857
ECP3153.961.042−0.9900.620
ECP4153.971.084−1.0730.697
Table 2. Common method bias analysis.
Table 2. Common method bias analysis.
ComponentInitial EigenvaluesExtraction Sums of Squared Loadings
Total% of VarianceCumulative %Total% of VarianceCumulative %
113.45534.50034.50013.45534.50034.500
23.1938.18742.6863.1938.18742.686
32.5426.51749.2032.5426.51749.203
42.0705.30754.5102.0705.30754.510
51.6644.26658.7761.6644.26658.776
61.6114.13162.9071.6114.13162.907
71.3583.48366.3901.3583.48366.390
81.1793.02269.4131.1793.02269.413
Extraction Method: Principal Component Analysis.
Table 3. Scale reliability testing.
Table 3. Scale reliability testing.
Construct ItemCorrected Item
–Total Correlation		Cronbach’s Alpha if Item DeletedCronbach’s α
Platform AI resourcesAI technology resourcesAIT10.8520.8990.925
AIT20.7230.924
AIT30.6710.923
AIT40.9270.883
AIT50.8590.898
AI human resourcesAIH10.6820.8550.876
AIH20.6210.869
AIH30.7690.834
AIH40.6440.864
AIH50.8190.822
Green value co-creationGreen co-productionGCP10.7210.8740.894
GCP20.6960.880
GCP30.7040.879
GCP40.7030.879
GCP50.9020.839
Green value-in-useGVIU10.7430.9060.918
GVIU20.7130.908
GVIU30.7530.905
GVIU40.7200.908
GVIU50.7220.908
GVIU60.7230.908
GVIU70.7130.909
GVIU80.7390.907
Firm performanceFirm performanceFP10.6580.8330.856
FP20.7230.806
FP30.6560.833
FP40.7590.790
Sustainable developmentEnvironmental performanceENP10.5910.7870.814
ENP20.6530.757
ENP30.6190.774
ENP40.6760.746
Social performanceSP10.7470.7810.848
SP20.6700.813
SP30.6510.821
SP40.6760.811
Economic performanceECP10.6920.7960.844
ECP20.6650.808
ECP30.6150.828
ECP40.7450.772
Total 0.957
Table 4. Structure validity testing.
Table 4. Structure validity testing.
Statistical Testχ2/dfRMSEAIFITLICFIGFIAGFI
Adaptation criteria1~3<0.08>0.9>0.9>0.9>0.8>0.8
Model values2.2640.0520.9290.9210.9280.8590.836
Whether it meets the standardsyesyesyesyesyesyesyes
Table 5. Convergent validity testing.
Table 5. Convergent validity testing.
ConstructItemStandard Load FactorAVECRCronbach’s α
AI technology resourcesAIT10.9110.7300.9300.925
AIT20.753
AIT30.696
AIT40.966
AIT50.913
AI human resourcesAIH10.7200.5970.8790.876
AIH20.669
AIH30.866
AIH40.676
AIH50.902
Green co-productionGCP10.7880.6500.9020.894
GCP20.786
GCP30.742
GCP40.704
GCP50.983
Green value-in-useGVIU10.7870.5850.9180.918
GVIU20.750
GVIU30.786
GVIU40.762
GVIU50.756
GVIU60.752
GVIU70.745
GVIU80.777
Environmental performanceENP10.6770.5260.8160.814
ENP20.737
ENP30.712
ENP40.773
Social performanceSP10.8560.5860.8490.848
SP20.758
SP30.712
SP40.727
Economic performanceECP10.7680.5790.8460.844
ECP20.754
ECP30.686
ECP40.830
Firm performanceFP10.6880.6020.8570.856
FP20.847
FP30.675
FP40.873
Table 6. Discriminant validity testing.
Table 6. Discriminant validity testing.
VariableAITAIHGCPGVIUECPENPSPFP
AIT0.854
AIH0.5580.773
GCP0.2890.3380.806
GVIU0.5300.4720.5340.765
ECP0.3350.2830.3820.5070.761
ENP0.3090.3350.3460.5640.5510.725
SP0.3890.3260.3350.4780.4870.6100.766
FP0.5030.5150.4680.5720.4780.4900.5340.776
AVE0.7300.5970.6500.5850.5790.5260.5860.602
Note: The diagonal is bolded to indicate the square root of the AVE.
Table 7. Correlation matrix and square root of AVE.
Table 7. Correlation matrix and square root of AVE.
AverageStandard Deviation1234567891011
1. AI technical resources3.631.0951
2. AI human resources3.461.1870.567 **1
3. Platform AI resources3.551.0100.875 **0.895 **1
4. Green co-production3.451.0460.314 **0.339 **0.369 **1
5. Green value-in-use3.451.1110.523 **0.446 **0.546 **0.512 **1
6. Green value co-creation3.450.9550.507 **0.462 **0.546 **0.788 **0.932 **1
7. Environmental performance3.920.8690.295 **0.321 **0.349 **0.319 **0.492 **0.487 **1
8. Social performance3.910.9050.348 **0.285 **0.356 **0.332 **0.421 **0.441 **0.512 **1
9. Economic performance3.940.8840.319 **0.256 **0.323 **0.336 **0.454 **0.467 **0.467 **0.415 **1
10. Sustainable development3.920.7100.400 **0.358 **0.427 **0.410 **0.568 **0.579 **0.819 **0.806 **0.782 **1
11. Firm performance3.571.1340.479 **0.453 **0.525 **0.477 **0.523 **0.576 **0.430 **0.471 **0.423 **0.551 **1
** p < 0.01 (2-tailed).
Table 8. Analysis of direct effects.
Table 8. Analysis of direct effects.
Firm PerformanceGreen Value Co-CreationFirm Performance
βtβtβt
Industry0.0641.6210.0691.8180.0270.707
Platform−0.021−0.535−0.013−0.347−0.002−0.057
Firm size−0.026−0.564−0.020−0.4440.0080.175
Firm age0.0300.6430.0220.484−0.018−0.393
Location0.0340.8680.0330.8570.0491.274
PAIRs0.52713.310 ***
PAIRs 0.57815.183 ***
GVC 0.56314.553 ***
R20.2820.3380.319
Adjusted R-squared0.2730.3290.310
F30.068 ***39.007 ***35.871 ***
*** p < 0.001.
Table 9. Analysis of the mediating role of GVC.
Table 9. Analysis of the mediating role of GVC.
Green Value Co-CreationFirm PerformanceFirm Performance
βtβtβt
Interaction term−0.023−0.5940.0641.6210.0742.038 *
Industry−0.012−0.298−0.021−0.535−0.016−0.452
Platform−0.020−0.437−0.026−0.564−0.018−0.422
Firm size0.0180.3930.0300.6430.0230.528
Firm age0.0090.2280.0340.8680.0310.848
PAIRs0.54513.932 ***0.52713.310 ***0.3026.995 ***
GVC 0.4139.563 ***
R20.3000.2820.402
Adjusted R-squared0.2900.2730.392
F32.727 ***30.068 ***43.916 ***
* p < 0.05, *** p < 0.001.
Table 10. Moderating role of sustainability in GVC and FP.
Table 10. Moderating role of sustainability in GVC and FP.
BSetpLLCIULCIR2F
Constant3.3800.17419.4140.0003.0383.7220.41941.205 ***
GVC0.3880.0586.7330.0000.2750.501
SD0.7960.1166.8550.0000.5681.024
Interaction term0.2180.0742.9360.0030.0720.363
Industry0.0530.0262.0220.0440.0020.105
Platform−0.0010.030−0.0240.981−0.0600.059
Firm size−0.0370.044−0.8290.408−0.1230.050
Firm age0.0300.0370.8180.414−0.0430.103
Location−0.0030.027−0.1030.918−0.0560.051
*** p < 0.001.
Table 11. Test results for the effects of predictor variables on the values of moderator variables.
Table 11. Test results for the effects of predictor variables on the values of moderator variables.
Sustainable DevelopmentBSetpLLCIULCI
−0.7100.2330.0932.5060.0130.0500.417
0.0000.3880.0586.7330.0000.2750.501
0.7100.5430.0599.1530.0000.4260.659
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Sun, Y.; Pongsakornrungsilp, S.; Pongsakornrungsilp, P.; Tonsakunthaweeteam, S.; Wongwaropakorn, W.; Chinchanachokchai, S. Platform AI Resources and Green Value Co-Creation: Paving the Way for Sustainable Firm Performance in the Digital Age. Sustainability 2025, 17, 8058. https://doi.org/10.3390/su17178058

AMA Style

Sun Y, Pongsakornrungsilp S, Pongsakornrungsilp P, Tonsakunthaweeteam S, Wongwaropakorn W, Chinchanachokchai S. Platform AI Resources and Green Value Co-Creation: Paving the Way for Sustainable Firm Performance in the Digital Age. Sustainability. 2025; 17(17):8058. https://doi.org/10.3390/su17178058

Chicago/Turabian Style

Sun, Yan, Siwarit Pongsakornrungsilp, Pimlapas Pongsakornrungsilp, Sasawalai Tonsakunthaweeteam, Wari Wongwaropakorn, and Sydney Chinchanachokchai. 2025. "Platform AI Resources and Green Value Co-Creation: Paving the Way for Sustainable Firm Performance in the Digital Age" Sustainability 17, no. 17: 8058. https://doi.org/10.3390/su17178058

APA Style

Sun, Y., Pongsakornrungsilp, S., Pongsakornrungsilp, P., Tonsakunthaweeteam, S., Wongwaropakorn, W., & Chinchanachokchai, S. (2025). Platform AI Resources and Green Value Co-Creation: Paving the Way for Sustainable Firm Performance in the Digital Age. Sustainability, 17(17), 8058. https://doi.org/10.3390/su17178058

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop