Sage Journals: Discover world-class research

Abstract

This study investigates the role of Artificial Intelligence (AI) in promoting environmental with in the manufacturing sector of small and medium enterprises (SMEs) in Pakistan. Grounded in Social Cognitive Theory (SCT), Resource-Based View (RBV), and Organizational Culture Theory (OC Theory), this research integrates AI adoption with Green Human Resource Management (GHRM) to explore how advanced technologies facilitate sustainable practices. Social Cognitive Theory (SCT) emphasizes the role of individual behavior in shaping organizational outcomes. In this study, it specifically highlights how AI-enabled tools can influence Green Employee Behavior (GEB), empowering employees to engage in eco-friendly practices that contribute to sustainability. Resource-Based View (RBV) suggests that organizations can achieve a competitive advantage by leveraging valuable, rare, and inimitable resources. This theory justifies the integration of AI adoption and sustainability-driven HR practices as strategic resources that SMEs can use to enhance their environmental performance. Organizational Culture Theory (OC Theory) underscores the importance of shared organizational values and norms in supporting the adoption of sustainability practices. It suggests that a strong pro-environmental culture enhances the acceptance and effectiveness of AI-driven GHRM practices. The study introduces GEB as a mediating factor, emphasizing the importance of employee-driven actions in translating AI-powered HR strategies into tangible environmental improvements. Additionally, Organizational Culture (OC) is assessed as a moderator, demonstrating how a pro-environmental culture strengthens the impact of AI and GHRM on sustainability outcomes. A quantitative research approach was employed, collecting data from 427 SMEs in Pakistan`s manufacturing sector. The PLS-SEM technique was used for empirical validation, confirming that AI adoption significantly enhances environmental sustainability through its influence on GHRM and GEB. Moreover, the results indicate that a strong pro-environmental culture amplifies these effects, reinforcing the role of OC as a critical moderator. This study advances theoretical discussions by integrating AI, GHRM, and sustainability within the SME context, offering a novel framework for understanding technology-driven sustainability initiatives. The findings provide actionable insights for SME leaders and policymakers, emphasizing the need for AI adoption, employee engagement, and the cultivation of sustainability-oriented organizational cultures. These insights contribute to the broader discourse on AI-driven sustainability strategies in emerging economies, providing a pathway for SMEs to achieve long-term environmental resilience.

Keywords

artificial intelligence green human resource management environmental sustainability green employee behavior organizational culture PLS-SEM SME manufacturing sector

Introduction

The global sustainability crisis has presented businesses with unprecedented challenges, compelling them to rethink their strategies and practices to mitigate environmental impacts. In this context, small and medium enterprises (SMEs) play a critical role due to their significant contributions to economic development, accounting for approximately 90% of all businesses and contributing 40% of GDP in Pakistan, where over 5 million SMEs dominate the business landscape (Matloob et al., 2023). However, these enterprises also contribute heavily to industrial pollution, inefficient energy consumption, and greenhouse gas emissions, underscoring the urgency of adopting sustainable practices. Addressing these challenges requires innovative approaches, and Artificial Intelligence (AI) has emerged as a transformative force in enabling sustainability, particularly in enhancing efficiency and reducing industrial environmental risks.

Despite the growing recognition of AI’s potential to mitigate these environmental challenges, its practical application in SMEs remains limited, particularly in terms of its impact on environmental sustainability. For instance, AI adoption is often more widely adopted in larger enterprises, which can better afford the technological infrastructure and investments required. However, some scholars argue that AI adoption presents a double-edged sword, where the initial high costs of implementation and potential resistance to change in smaller organizations may limit its effectiveness (Aliyah et al., 2023). At the same time, the potential benefits of AI adoption for SMEs such as optimizing resource usage, reducing waste, and lowering energy consumption are well-documented. For example, a recent study by Qasim et al. (2025) demonstrated how AI-powered systems in SMEs in Pakistan’s manufacturing sector helped reduce energy consumption by 25% in 1 year. This example illustrates how AI adoption can support environmental sustainability while improving operational efficiency in resource-constrained environments.

AI technologies, such as machine learning, natural language processing, and data analytics, offer SMEs the tools to optimize resource utilization, streamline operations, and enhance their environmental sustainability (Qasim et al., 2025). AI-driven Internet of Things (IoT) systems, for instance, can monitor and reduce energy consumption in real-time, significantly improving energy efficiency in manufacturing operations. Studies suggest that these technologies can reduce energy waste by up to 30% in developing countries (Aliyah et al., 2023; Schwäke et al., 2024). These capabilities enable SMEs to align their business processes with sustainability goals, addressing environmental concerns while maintaining operational efficiency. However, AI adoption remains challenging for SMEs in developing economies like Pakistan due to limited resources, lack of digital infrastructure, and resistance to change (Hongyun et al., 2023).

Previous studies have explored AI in organizations, but there is limited research on how AI-driven HRM practices, particularly Green Human Resource Management (GHRM) and Green Employee Behavior (GEB), contribute to environmental sustainability in SMEs, especially in developing countries like Pakistan. Existing studies have primarily focused on large organizations or examined these elements in isolation, leaving a significant research gap regarding the intersection of AI, GHRM, GEB, and Organizational Culture (OC) in the context of SMEs, particularly in emerging economies like Pakistan.

This research aims to fill these gaps by proposing a novel conceptual framework that integrates AI, GHRM, GEB, and OC, offering new insights into their combined impact on environmental sustainability in SMEs. Theoretically, this study introduces GEB as a mediator and OC as a moderator, providing fresh perspectives on how these elements interact to enhance sustainability in SMEs. Practically, this study provides actionable recommendations for SME leaders and policymakers on how to effectively adopt AI-driven sustainability practices and integrate green HRM practices to achieve long-term environmental and energy-related sustainability.

SMEs in Pakistan, while contributing significantly to the economy, are major contributors to environmental degradation due to inefficient resource usage and unsustainable practices. Despite the growing recognition of AI’s potential to improve efficiency and reduce environmental harm, the adoption of AI-driven sustainability practices remains limited in Pakistan’s SME sector. The energy-intensive manufacturing sector in Pakistan is a major contributor to pollution and resource depletion, and it has received limited attention in studies on AI and sustainability. Addressing these gaps is essential for understanding how SMEs in Pakistan can leverage AI-driven GHRM practices to foster GEB and achieve environmental and energy-related sustainability objectives.

One promising avenue for leveraging AI adoption in sustainability is through Green Human Resource Management (GHRM). GHRM integrates environmental principles into HR practices, including green recruitment, sustainability-focused training, and eco-conscious performance evaluations (Sharma et al., 2022). AI enhances these practices by enabling efficient talent acquisition aligned with sustainability goals, automating training modules, and incorporating environmental metrics into performance appraisals (Aydin et al., 2024). For example, AI-powered systems can help identify employees who are most likely to embrace eco-friendly behaviors and tailor training to enhance their environmental contributions. However, the success of AI-driven GHRM initiatives often depends on individual and organizational factors. Employees’ willingness to adopt green practices and the presence of a supportive organizational culture are critical for translating AI-enabled strategies into meaningful sustainability outcomes (Sulaiman et al., 2016).

Research shows that GEB plays a pivotal role in this context, as it involves proactive actions by employees to support environmental objectives. AI-powered tools such as gamified training platforms can motivate employees to engage in eco-friendly practices, such as energy conservation and waste reduction (Shafaei et al., 2020). However, fostering GEB requires more than just technology; A supportive OC is equally essential. Organizational Culture Theory (OCT) emphasizes the role of shared values, beliefs, and norms in shaping employees’ attitudes toward sustainability. OC encompasses the shared values, beliefs, and norms that shape employees’ attitudes and behaviors toward sustainability. Organizations with a pro-environmental culture are more likely to successfully implement AI-driven GHRM practices, as employees are more inclined to embrace green initiatives in such environments (Sheikh et al., 2024). For example, a study by Wang (2019) highlighted how organizations with strong sustainability values have higher rates of employee participation in environmental programs. Despite the growing recognition of AI, GHRM, and GEB as critical components of sustainability, their intersection remains underexplored in the SME context, particularly in developing economies. Existing studies have primarily focused on large organizations or examined these elements in isolation, leaving a significant research gap. Moreover, the energy-intensive manufacturing sector in Pakistan, a major contributor to pollution and resource depletion has received limited attention in studies on AI and sustainability. Addressing these gaps is essential for understanding how SMEs can leverage AI-driven GHRM practices to foster GEB and achieve environmental and energy-related sustainability objectives.

This study aims to fill these gaps by investigating the relationships between AI adoption, GHRM, GEB, and OC in promoting environmental sustainability within SMEs. Specifically, it addresses the following objectives:

Assess the direct impact of AI on ES in SMEs.

Evaluate the influence of AI on GHRM in SMEs.

Investigate the mediating role of GEB in the relationship between AI and ES.

Examine the moderating effect of OC on the relationship between AI and ES.

This study examines both organizational-level variables (such as AI adoption, GHRM, and OC) and individual-level variables (such as GEB). By integrating these levels, the study adopts a cross-level approach, offering a comprehensive understanding of how AI-driven HRM practices influence environmental sustainability in SMEs.

The theoretical foundation of this study draws on the RBV, SCT, and OCT. RBV suggests that organizations can achieve a competitive advantage by leveraging valuable, rare, and inimitable resources. This theory justifies the integration of AI and sustainability-driven HR practices as strategic resources that SMEs can use to enhance their environmental performance (Garg et al., 2018). SCT emphasizes the role of individual behaviors in shaping organizational outcomes, particularly how AI-enabled tools can influence GEB, empowering employees to engage in eco-friendly practices that contribute to sustainability (Marks & Bandura, 2002). OC Theory underscores the importance of shared organizational values and norms in supporting the adoption of sustainability practices. It suggests that a strong pro-environmental culture enhances the acceptance and effectiveness of AI-driven GHRM practices (Wang, 2019).

Methodologically, this study employs PLS-SEM to test the proposed relationships. PLS-SEM is particularly suited for analyzing complex models with multiple constructs and mediating or moderating effects (Hair et al., 2022). By integrating quantitative and theoretical insights, this research provides a nuanced understanding of how SMEs can harness AI-driven GHRM practices to foster sustainability.

The findings of this study contribute significantly to both theory and practice. Theoretically, it advances the literature on AI adoption, GHRM, and sustainability by introducing GEB as a mediator and OC as a moderator, offering new insights into their interactions. Practically, it provides actionable recommendations for SME leaders and policymakers. This research emphasizes the need to invest in AI-powered HR tools, foster a pro-environmental culture, and engage employees in sustainability initiatives to maximize the benefits of AI integration. These insights are particularly relevant for SMEs in resource-constrained environments, where leveraging technology for sustainability can also enhance competitiveness and resilience.

This study is organized as follows: the subsequent section reviews the relevant literature and theoretical frameworks, followed by a detailed explanation of the research methodology. The results of the PLS-SEM analysis are then presented, highlighting the direct, indirect, and moderated effects. Finally, the discussion and conclusion sections outline the theoretical and practical implications of the findings, identify limitations, and suggest directions for future research.

By addressing the complex interplay of AI adoption, GHRM, GEB, and OC in the SME context, this study provides a comprehensive framework for understanding how technology and human dynamics can jointly drive environmental sustainability and energy efficiency. It also underscores the importance of cultural and behavioral factors in amplifying the impact of AI-driven strategies, offering valuable insights for academics, practitioners, and policymakers seeking to foster sustainability in SMEs.

Literature Review

Artificial Intelligence Adoption and Environmental Sustainability

AI Adoption is a key enabler of sustainable environments by optimizing organizational resources and enhancing efficiency. Similar to data analytics, predictive analysis, and real time monitoring, AI adoption helps organizations optimize their energy consumption, reduce environmental waste, and lower their carbon footprint. It also aids in the creation of sustainable products and services (Sipola et al., 2023). This study examines both organizational-level variables (such as AI adoption, GHRM, and OC) and individual-level variables (such as GEB), recognizing the dynamic interplay between organizational practices and individual behaviors in driving environmental sustainability within SMEs.

For SMEs, the practical application of AI adoption to drive environmental sustainability remains limited. However, its potential impact on sustainability is becoming increasingly evident. AI adoption has been shown to assist SMEs in reducing waste, improving energy efficiency, and adapting operations to sustainability-focused models. Specifically, AI enables organizations to optimize resource usage, reduce waste, and improve energy conservation. For instance, AI-driven systems have reduce energy consumption by 25% within 1 year in Pakistan’s manufacturing sector (Qasim et al., 2025; Waqas et al., 2021). This case illustrates AI’s practical impact on sustainability goals, aligning with the concept of AI adoption as a transformative force for improving environmental performance in resource-constrained enterprises. While AI’s potential in sustainable services and products and energy conservation is significant, these are broader concepts. In line with the study’s definition of environmental sustainability, the evidence supporting H1 AI adoption positively influences ES in SMEs now emphasizes AI’s role in operational efficiency, resource optimization, and waste reduction. This ensures the evidence aligns with the study’s specific focus on environmental sustainability.

Despite its potential, AI adoption in SMEs often faces barriers such as limited resources, lack of digital infrastructure, and resistance to change, especially in developing countries like Pakistan (Hongyun et al., 2023). Some scholars argue that AI presents a double-edged sword while it offers environmental benefits, high implementation costs and technological challenges can outweigh these benefits, especially in smaller firms (Aliyah et al., 2023). These challenges highlight the need for further research on how AI can be integrated into SMEs for sustainable operations.

Thus, AI adoption is recognized as a significant driver of both environmental and energy sustainability in SMEs, particularly in resource-intensive production industries, where the need for efficiency and sustainability is paramount. AI-driven systems enable SMEs to adopt more energy-efficient production methods reduce overall energy demand, and monitor the environmental impact of their products, helping businesses maintain compliance with sustainability standards. These capabilities position AI adoption as a key driver of both environmental and energy sustainability in SMEs, especially in the resource-intensive industries.

H1: AI adoption positively influences ES in SMEs.

AI Adoption and Green HRM

AI adoption is transforming GHRM by integrating advanced technologies into HR processes, making organizations more effective in managing human capital sustainably (Soomro et al., 2024). AI tools also improve green recruitment, helping businesses identify candidates with a sustainability-oriented mindset and relevant experience in environmentally responsible practices (Alzyoud, 2021). Additionally, AI adoption enhances green training programs, equipping employees with the necessary knowledge and skills to support sustainability initiatives (Alzoubi & Mishra, 2024). This study considers both organizational-level variables (such as AI adoption and GHRM) and individual-level variables (such as GEB), to better understand how AI-driven GHRM practices can influence environmental sustainability at both the organizational level and the individual level.

AI-driven performance management systems ensure that employees are assessed based on their contribution to environmental goals, promoting sustainability as a business performance indicator (Sohu et al., 2025). Furthermore, AI-powered HR systems improve workforce planning, enhancing resource efficiency and reducing energy waste, thus strengthening the environmental impact of GHRM. These technological innovations not only improve HR processes but also strengthen GHRM implementation in SMEs, contributing to energy-efficient and sustainable business practices. Despite these advancements, the success of AI-driven GHRM initiatives often depends on individual and organizational factors. Employees’ willingness to adopt green practices and the presence of a supportive organizational culture are critical for translating AI-enabled strategies into meaningful sustainability outcomes (Sulaiman et al., 2016). This is consistent with OC Theory, which emphasizes how cultural values shape employees’ attitudes toward sustainability. Thus, the intersection of AI and GHRM offers significant potential to enhance sustainability outcomes in SMEs.

H2: AI adoption positively impacts Green HRM in SMEs.

GHRM and Environmental Sustainability

GHRM integrates environmental objectives into HR policies and practices, creating a framework for fostering eco-friendly behaviors and energy-conscious workplace practices (Shaikh et al., 2024). Key GHRM practices such as green recruitment, sustainability-focused training and environmentally responsible performance evaluation, ensure that employees contribute to environmental goals. Green recruitment ensures that new hires possess the skills and commitment necessary to contribute to sustainability objectives, while green training programs equip employees with the knowledge about energy conservation, efficient resource utilization, and eco-friendly business strategies (Sharma et al., 2022).These practices foster employee involvement in sustainability, translating organizational goals into practical, measurable outcomes.

For instance, green recruitment ensures that new hires are aligned with sustainability objectives, while green training empowers employees with knowledge on energy conservation, efficient resource use, and environmentally conscious practices. AI-powered performance management systems help integrate environmental Key Performance Indicators (KPIs), motivating employees to actively participate in sustainable business operations. These systems enable the incorporation of environmental KPIs into performance assessments, ensuring employees contribute to energy-efficient practices and sustainability goals (Sun et al., 2022). In the SME manufacturing context, where high energy consumption and environmental degradation are significant challenges, GHRM serves as a critical mechanism to mitigate environmental impact and drive sustainable and energy-efficient practices.

H3: Green HRM positively contributes to ES in SMEs.

AI Adaption Positively Contributes GEB

Artificial intelligence adoption plays a pivotal role in shaping employees’ pro-environmental actions by fostering awareness, motivation, and behavioral reinforcement toward sustainability. Through intelligent systems such as real-time monitoring, predictive analytics, and gamified learning platforms, AI empowers employees to identify, evaluate, and modify environmentally impactful behaviors (Shafaei et al., 2020). AI-enabled HR systems personalize green training, provide instant feedback, and reward eco-friendly performance, thereby enhancing employees’ environmental self-efficacy and engagement in sustainability initiatives (Aydin et al., 2024). This interaction aligns with social cognitive theory, which posits that behavior is shaped through observational learning, feedback, and self-regulation (Marks & Bandura, 2002). In this context, AI functions as a cognitive and behavioral enabler, reinforcing employees’ green attitudes and translating them into concrete environmental actions.

H4: AI adoption positively influences green employee behavior in SMEs.

GEB Positively Contributes to ES

Green employee behavior represents employees’ voluntary and task-related actions that contribute to reducing the organization’s ecological footprint and achieving environmental objectives (Shafaei et al., 2020). Employees who actively engage in energy conservation, waste reduction, and sustainable resources use directly enhance environmental sustainability outcomes within SMEs (Shaikh et al., 2024). Prior studies affirm that GEB serves as a micro-level driver translating organizational sustainability policies into tangible environmental improvements (Sun et al., 2022). When employees consistently demonstrate pro-environmental commitment such as responsible energy consumption and eco-innovation, the cumulative impact significantly strengthens organizational sustainability performance. Guided by social cognitive theory, such behaviors stem from learned environmental values and self-regulatory mechanisms reinforced by organizational systems (Marks & Bandura, 2002).

H5: GEB positively influences environmental sustainability in SMEs.

Mediating Role of GHRM and GEB

AI adoption also indirectly enhances environmental sustainability through the mediating role of GHRM. By embedding AI technologies into HR functions, organizations can automate green recruitment, sustainability training, and environmental performance evaluations, aligning human capital practices with ecological objectives (Aydin et al., 2024; Soomro et al., 2024). Such AI-driven GHRM systems strengthen employees’ environmental awareness and participation in sustainability initiatives, thereby translating technological integration into improved environmental outcomes. This synergy underscores that AI adoption fosters environmental sustainability not only through behavioral change but also by institutionalizing green HR practices within organizational processes (Sulaiman et al., 2016).

Green employee behavior acts as a crucial mediator, translating organizational sustainability strategies into tangible environmental and energy-efficient outcomes. GEB involves employees’ proactive engagement in sustainable actions, such as energy conservation, waste reduction, and participation in corporate environmental programs (Shafaei et al., 2020).

AI-powered systems, such as gamified training platforms, can enhance employees’ pro-environmental attitudes by educating them about the negative environmental impacts of certain behaviors and providing energy-efficient alternatives. Additionally, AI-based performance tracking systems incentivize positive sustainability behaviors, reinforcing long-term engagement in green initiatives. For example, an AI adoption system could track and reward employees for reducing energy consumption or promoting waste reduction behaviors in the workplace.

Moreover, AI’s ability to personalize training and track progress ensures that employees are continuously motivated to participate in sustainability initiatives, making GEB a key mediator between AI adoption and ES. These systems not only improve employee engagement but also drive measurable environmental improvements and enhanced energy conservation, which are core elements of environmental sustainability in SMEs.

H6: The impact of AI adoption on ES is mediated through GHRM.

H7: The impact of AI adoption on ES is mediated through GEB.

Moderation of Organizational Culture

Organizational culture (OC) serves as a moderating factor influencing the effectiveness of AI adoption and GHRM in achieving environmental sustainability and energy efficiency. A strong pro-environmental culture fosters employee buy-in for green initiatives, amplifying the positive effects of sustainability strategies (Sheikh et al., 2024). A supportive OC enhances the adoption of AI-driven GHRM practices, making employees more receptive to green HR policies, sustainability training and energy-saving initiatives (Manika et al., 2013). In energy-intensive industries, a strong pro-sustainability culture ensures that AI-driven energy conservation strategies are successfully implemented.

However, past research has shown inconsistencies regarding the role of organizational culture in moderating the effectiveness of AI and GHRM practices. Some studies suggest that organizational culture significantly influences the adoption of sustainability practices, but others argue that cultural resistance can weaken the effectiveness of sustainability strategies (Aliyah et al., 2023; Chand et al., 2024). Specifically, organizations with weak or conflicting environmental values face greater challenges in implementing sustainability-focused initiatives, especially those driven by technology like AI. In contrast, organizations with strong pro-environmental values experience smoother adoption and higher success rates in AI-driven energy conservation strategies (Lee et al., 2019).

Therefore, OC plays a crucial moderating role, as a strong environmental culture is essential for the successful implementation of AI-driven sustainability initiatives. The presence of a pro-sustainability culture increases employee engagement, supports AI adoption, and enhances the effectiveness of GHRM practices, ultimately maximizing environmental sustainability outcomes in SMEs.

H8: OC moderates the relationship between GHRM and ES, strengthening the positive impact.

Methods

Sample and Data Collection

This research adopted a quantitative research methodology to examine the relationship between the proposed factors of AI, GHRM and ES, where GEB playing a mediating role and OC acting as the moderating factor. Following the positivist philosophy this approach enabled the examination of association between the study constructs while ensuring empirical rigor.The research context was focused on the SME manufacturing sector, a key contributor to industrial sustainability efforts and economic growth. This sector was chosen due to its significant environmental footprint, making it crucial for implementing AI-driven sustainability initiatives. Figure 1 illustrates the conceptual model of current study, showing the relations between endogenous and exogenous Constructs.

Figure 1.

Conceptual framework.

Given the impracticality of obtaining a complete list of all SMEs, a nonprobability convenience sampling method was adopted. This method is effective in maintaining data collection rigor while ensuring feasibility. The selection criteria for SMEs included operational organizations that have been in existence for over 2 years, with respondents being SME owners, managers, or executives who have at least 2 years of managerial experience.

The survey targeted SMEs in the manufacturing sector within Pakistan, specifically focusing on companies with fewer than 250 employees. Respondents were SME owners, CEOs, executives, and managers who were directly involved in sustainability practices and AI adoption in their organizations. The enterprises surveyed were operational for over 2 years, ensuring they fit within the SME classification. This context ensures the reliability of the data, as both the respondents’ roles and the enterprise characteristics are aligned with the research objectives.

To collect data, an online structured questionnaire was used. The questionnaire was first piloted and cross-validated by domain experts and initial respondents. A pilot test was conducted with 30 participants from the target population (SME owners, CEOs, executives, and managers), and the feedback was used to revise the final version of the questionnaire.

Data collection occurred from June to September 2024, targeting SMEs in densely populated manufacturing hubs, due to their high environmental impact and potential for AI-driven sustainability adoption. Ethical research guidelines were strictly followed, including obtaining informed consent from participants, ensuring anonymity, and providing an official research brief outlining the study’s purpose and implications for sustainable practices. A total of 1,157 questionnaires were distributed, and 541 valid responses were received, yielding a 46.76% response rate. After removing 114 incomplete or invalid responses, 427 completed questionnaires were used for analysis. This sample size meets the minimum requirements for PLS-SEM analysis, as recommended by Hair et al. (2022). minimum sample size calculation was performed, considering the complexity of the model (with five constructs and multiple variables).

Based on these calculations, the required minimum sample size was 250 participants. The final sample size of 427 responses exceeds the required minimum, ensuring statistical power and robustness for analyzing the structural relationships within the model. Among the 427 responses, 287 respondents were male and 140 were female, with further demographic details presented in Table 1.

Table 1.

Demographic Profile of Respondents.

Demographic variables	Categories	Frequency (n)	Percentage (%)
Gender	Male	287	67.2
	Female	140	32.8
Educational qualifications	Graduate	275	64.4
	Postgraduate	125	29.3
	PhD	27	6.3
Years of experience	5–10 years	235	55.0
	11–15 years	130	30.4
	16–20 years	45	10.5
	21–25 years	17	4.1

Variables and Measurement

This study encompasses five key constructs, including two independent variables, two mediators, one moderator, and one dependent variable. While scales were adapted from validated sources, some items were refined after pilot testing (see Table 2). The scales were selected based on their relevance to the theories (SCT, RBV, OC Theory) that underpin the study and the established reliability and validity of previous studies. The complete list of measurement scales used in this study are provided in Appendix 1.

Table 2.

Questionnaire Items After Pilot Testing.

Item	Source	Amendments
AIA1: Our organization uses AI tools to automate routine and repetitive tasks	Garg et al. (2018)	No changes
GHRM1: Our enterprise provides adequate training to promote environmental management as a core organizational value	Kim et al. (2023)	No changes
GEB1: I actively participate in environmental initiatives organized by my workplace	Shafaei et al. (2020)	Revised wording for clarity after pilot feedback
OC1: Preserving the environment is a central corporate value in our firm	Wang (2019)	Revised to include more specific actions related to sustainability practices
ES1: Our SME adopts sustainable resource management practices to minimize waste and optimize efficiency	Carrasco-Carvajal et al. (2023)	No changes

Artificial Intelligence Adoption (AIA) was measured using six items adapted from (Garg et al., 2018), capturing AI’s role in automation, decision-making, and innovation. A sample question is: “Our organization uses AI tools to automate routine and repetitive tasks.”

Green Human Resource Management (GHRM) was assessed with six items adapted from Kim et al. (2023), evaluating green recruitment, training, performance evaluation, and alignment with environmental goals. A sample question is: “Our enterprise provides adequate training to promote environmental management as a core organizational value.”

Green Employee Behavior (GEB) was measured using five items adapted from Shafaei et al. (2020), reflecting employees’ participation in sustainability initiatives. A sample item is: “I actively participate in environmental initiatives organized by my workplace.”

Organizational Culture (OC) was evaluated using six items adapted from Wang (2019), capturing pro-environmental values and practices within an organization. The presence of a strong sustainability-oriented culture influences the adoption of green practices. A sample question is: “Preserving the environment is a central corporate value in our firm.”

Environmental Sustainability was assessed using five items adapted from Carrasco-Carvajal et al. (2023) evaluating sustainability practices such as resource management and eco-friendly operations. A sample item is: “Our SME adopts sustainable resource management practices to minimize waste and optimize efficiency.”

Pilot Test and Questionnaire Modifications

To ensure the reliability and validity of the questionnaire, a pilot test was conducted with 30 participants from the target population (SME owners, CEOs, executives, and managers). Feedback from the pilot test was used to revise the final version of the questionnaire. Below is a Table 2 detailing the questionnaire items, their sources, and any modifications made after the pilot test.

Measurement Scale

All constructs were rated on a 5-point Likert scale (1 = “Strongly Disagree” to 5 = “Strongly Agree”). Reliability and validity were assessed using Cronbach’s alpha (Constructs values are greater than .7), composite reliability (Constructs values greater than .7), and average variance extracted (constructs values greater than .5), All constructs met the recommended thresholds (Hair et al., 2019), confirming their suitability for further analysis. The selection of these variables ensures a comprehensive evaluation of AI-driven sustainability strategies in SMEs, particularly in energy-intensive sectors. By using scales that have been validated in the literature, we ensure that the constructs effectively capture the relevant dimensions for evaluating AI-driven sustainability in SMEs.

Data Analysis

The PLS-SEM method was chosen for this study due to its flexibility and usefulness in analyzing complex models that incorporate multiple constructs and interaction between construct levels (Hair et al., 2019). PLS-SEM is well suited to handle intricate theoretical models, making it ideal for validating the relationships between the proposed constructs in this study.

The method focuses on two aspects: the measurement model (which assesses the reliability and validity of the constructs) and the structural model (which tests the relationships between the variables). The measurement model ensures that the constructs used in the study are valid and reliable, while the structural model examines the hypothesized causal relationships between the variables.

To explore the causal relationships between the independent variables (AI adoption and GHRM), the mediator (GEB), the moderator (OC), and the dependent variable (ES), PLS-SEM offers significant advantages. This dual focus enables a comprehensive evaluation of the hypothesized relationships, ensuring that both the constructs’ validity and the relationships between them are rigorously tested.

In addition, to ensure more robust validation of the mediation and moderation effects, the Bootstrap method (Hair et al., 2022) was incorporated. This additional step improves the accuracy and reliability of the results by providing more robust estimates for indirect effects. By combining PLS-SEM with the Bootstrap method, this study strengthens the validity of its findings, offering empirical insights into how AI-driven GHRM practices, OC, and GEB contribute to environmental sustainability in SMEs.

The use of these advanced techniques ensures that the theoretical contributions of the study are rigorously validated, providing actionable insights for SME leaders and policymakers working to integrate AI into sustainable business practices. By employing PLS-SEM, the study enhances the credibility of its results, making them highly relevant for SME leaders and policymakers aiming to integrate AI into sustainable business practices.

Results

Testing for Convergent and Discriminant Validity

Convergent and discriminant validity findings are summarized in Table 3. The outer loadings in the measurement model ranged from .716 to .878, which was good because all the loadings are above the recommended threshold of .7. These results show high correlations between the measured variables and their corresponding theoretical constructs (Hair et al., 2019). All CR values obtained were above .7, the minimum acceptable value, thereby confirming the internal consistency of the constructs; the range was .894 to .945. Likewise, the average variance extracted was found to be within the acceptable range for all constructs with values ranging .628 to .742 confirming the convergent validity were above .5. The reliability test showed that Cronbach’s alpha values for all constructs lies between .852 and .931. These results support the reliability and validity of the above-mentioned construct in this study.

Table 3.

Reliability and Validity.

Construct reliability and validity
Items	Outer loadings	Constructs	Alpha	CR	AVE	Outer VIF
AIA1	.806	Artificial intelligence	.898	.921	.661	2.194
AIA2	.823					2.293
AIA3	.826					2.439
AIA4	.799					2.152
AIA5	.816					2.287
AIA6	.809					2.347
ES1	.716	Environmental sustainability	.863	.901	.647	1.537
ES2	.805					2.467
ES3	.841					2.612
ES4	.817					2.518
ES5	.837					2.619
GEB1	.821	Green employee behavior	.852	.894	.628	2.313
GEB2	.798					1.875
GEB3	.765					1.883
GEB4	.799					2.073
GEB5	.777					1.993
GHRM1	.875	Green human resource management	.931	.945	.742	3.005
GHRM2	.878					3.913
GHRM3	.838					2.634
GHRM4	.865					2.833
GHRM5	.853					2.789
GHRM6	.860					3.412
OC1	.794	Organizational culture	.885	.912	.634	2.156
OC2	.795					2.418
OC3	.817					2.875
OC4	.758					1.963
OC5	.790					2.339
OC6	.823					2.721

Discriminant Validity

The discriminant validity was analyzed using the Heterotrait–Monotrait Ratio (HTMT), as presented below Table 4. HTMT values below .85 (or .90 in exploratory research) is considered acceptable discriminant validity, which proves that the constructs are different from each other (Hair et al., 2017).

Table 4.

Discriminant Validity.

Heterotrait–monotrait ratio (HTMT)—Matrix	AI	ES	GEB	GHRM
Environmental sustainability	0.771
Green employee behavior	0.693	0.781
Green human resource management	0.669	0.828	0.596
Organizational culture	0.684	0.869	0.654	0.768

Inner Model VIF and f-Square

Variance inflation factor (VIF) was used to measure multicollinearity in the inner model. According to (Kock, 2015), VIF values below 5 indicate no significant multicollinearity concerns, ensuring reliable estimations. As shown in Table 5, all constructs exhibit VIF values below the critical threshold. For instance, AI shows VIF values of 2.056 for ES, and 1.000 for both GEB and GHRM, confirming minimal collinearity. Similarly, OC has a VIF of 2.355 for ES, supporting its independent contribution. These results confirm that the predictors are distinct and contribute uniquely to the variance of their respective dependent variables, ensuring robust, and unbiased estimations in the inner model.

Table 5.

Inner Model VIF and f-Square.

Inner VIF	ES	GEB	GHRM
Green employee behavior	1.790
Green human resource management	2.266
Organizational culture	2.355
f-square
Artificial intelligence adoption	0.046	0.590	0.610
Green employee behavior	0.119
Green human resource management	0.163
Organizational culture	0.185

Model Fit, Predictive Validity, and R-Square

The Standardized Root Mean Square Residual (SRMR) for both the saturated model (.063) and the estimated model (.068) falls below the threshold of .10, indicating a well-fitting model. An SRMR value closer to 0 signifies better fit, confirming that the model effectively captures the variance-covariance structure of the data. Additionally, the Normed Fit Index (NFI) values for the saturated model (.806) and the estimated model (.803) approach the recommended threshold of .90, suggesting an acceptable model fit, particularly for exploratory research involving complex constructs.

The PLS Q² Predict metric assesses predictive validity. Q² values greater than 0 indicate acceptable predictive power, while RMSE evaluates prediction accuracy (Hair, Gabriel, et al., 2019). As shown in Table 6, ES has a high Q² Predict value of .631 and a low RMSE of .610 revealing its capacity for accurate and consistent prediction (Dolce et al., 2017). Likewise, Q² Predict for GEB and GHRM are .366 and .376 for RMSE at .801 and .794, respectively. Such results underline the measure of the model’s reliability and its potential for forecasting future results.

Table 6.

Model Fit, Predictive Validity, and R-Square.

Model fit, predictive validity, and R-Square.	R-square	R-square adjusted
R-square
Environmental sustainability	0.747	0.744
Green employee behavior	0.371	0.370
Green human resource management	0.379	0.377
	Saturated model	Estimated model
Model fit
SRMR	0.063	0.068
NFI	0.806	0.803
	Q2 Predict	RMSE
PLS predict
Environmental sustainability	0.631	0.610
Green employee behavior	0.366	0.801
Green human resource management	0.376	0.794

R-Square (explained variance) values represent the proportion of variance in dependent variables explained by independent variables, with Adjusted R-Square accounting for model complexity (Gao, 2023). For ES, R² (.747) and Adjusted R² (.744) reveal that 74.7% of the variance is explained by the predictors, reflecting strong explanatory power. Similarly, for GEB, R² (.371) and Adjusted R² (.370) show that 37.1% of the variance is explained by the model, while for GHRM, R² (.379) and Adjusted R² (.377) indicate that 37.9% of the variance is explained. These results underscore the relevance of the predictors in explaining and forecasting outcomes. For further details, please refer to Table 6.

Hypothesis Testing and Mediation Effects

Structural Equation Model (SEM)

The results from PLS-SEM confirmed the relationships among the variables, as shown in Figure 2 and summarized in Table 7. AI had a significant direct effect on GEB, GHRM, and ES. Specifically, AI positively influenced GEB (β = .609, p < .000), GHRM (β = .616, p < .000), and ES (β = .154, p < .000). Additionally, GEB and GHRM both positively influenced ES. The R² values indicated that AI, GHRM, and OC explained 37.1%, 37.9%, and 74.7% of the variance in GEB, GHRM, and ES, respectively, demonstrating strong explanatory power.

Figure 2.

Structural model.

Table 7.

Direct Path Hypothesis.

Direct effect and moderation	Path	Mean	SD	T-value	p-Values	Decision
Artificial intelligence adoption → green employee behavior	0.609	0.609	0.031	19.918	.000	Accepted
Artificial intelligence adoption → green human resource management	0.616	0.616	0.028	21.748	.000	Accepted
Artificial intelligence adoption → environmental sustainability	0.154	0.154	0.035	4.453	.000	Accepted
Green employee behavior → environmental sustainability	0.233	0.232	0.036	6.553	.000	Accepted
Green human resource management → environmental sustainability	0.306	0.307	0.037	8.226	.000	Accepted
Organizational culture × green human resource management → environmental sustainability	0.082	0.082	0.020	4.009	.000	Accepted
Specific indirect effect
Artificial intelligence adoption → green human resource management → environmental sustainability	0.188	0.189	0.025	7.451	.000	Accepted
Artificial intelligence adoption→ green employee behavior → environmental sustainability	0.142	0.141	0.023	6.156	.000	Accepted

Mediation and Moderation Effects

The mediation effects in this study were analyzed to explore the indirect relationships among AI, GEB, GHRM, OC, and ES. Partial mediation was observed across all hypotheses involving indirect effects, emphasizing the critical role of mediating constructs. The results show that AI indirectly influences ES through GEB (H2), with a significant indirect effect (β = .142, p < .001), as well as through GHRM (H3), with a significant indirect effect (β = .188, p < .001), highlighting the importance of employee behavior and HR practices in translating AI capabilities into enhanced environmental outcomes (Tenenhaus et al., 2005).

The moderating role of OC in the relationship between GHRM and ES (β = .082, p < .001) was also confirmed, demonstrating how OC amplifies the positive effects of green HR on ES. Additionally, the parallel mediation pathway reveals that AI impacts ES through GHRM and GEB, further highlighting the need to integrate technological advancements with organizational strategies to achieve sustainability goals (Shmueli et al., 2016).

Table 7 summarizes the direct and indirect effects, demonstrating significant results for all hypotheses. The robust confidence intervals and significant p-values validate the model’s ability to explain the relationships among the constructs.

Discussion

This study investigated how AIA adoption, GHRM, GEB, and OC interact to enhance ES in manufacturing SMEs. By testing five hypotheses (H1–H5), the results provide robust empirical support for the theorized model, revealing that AI adoption positively affects GHRM and ES, that GHRM in turn contributes significantly to ES, that GEB mediates the pathway from AI to ES, and that OC moderates the relationship between GHRM and ES. Together, these findings illuminate how technology, people, and culture converge to create sustainable organizations in emerging economies. Below, the discussion elaborates on each hypothesis direct, mediating, and moderating effects by interpreting the findings in light of past studies, relevant theories, and contextual realities of SMEs.

Direct Effects

H1: AI Adoption → Environmental Sustainability

The results strongly support H1, confirming that AI adoption exerts a positive and significant effect on environmental sustainability in SMEs. This finding corroborates prior studies (Ababneh et al., 2021; Akter et al., 2022; Sipola et al., 2023) which emphasize AI’s potential to minimize waste, optimize resource utilization, and enhance operational efficiency through data-driven decision making. AI technologies such as predictive analytics, machine learning, and Internet of Things (IoT) sensors enable SMEs to monitor energy use in real time, identify inefficiencies, and reduce emissions across production processes.

This finding can be understood through the RBV: AI constitutes a strategic and inimitable resource that allows firms to reconfigure operations and achieve sustainable competitive advantage. In SMEs, which often operate under financial and infrastructural constraints, AI offers an affordable mechanism for energy and waste optimization without large-scale infrastructural overhauls. The results also reflect the diffusion of Industry 4.0 technologies within emerging markets, suggesting that digital transformation is not limited to large corporations but increasingly drives environmental performance in smaller firms.

Interestingly, the strength of the AI–ES relationship may also reflect contextual factors unique to Pakistan’s manufacturing SMEs, where energy inefficiency and environmental degradation remain pressing challenges. By adopting AI tools, these firms can leapfrog traditional inefficiencies and align their operations with global sustainability standards. Thus, AI adoption functions as a technological equalizer, empowering SMEs to compete on both ecological and economic dimensions.

H2: AI Adoption → Green Human Resource Management

Hypothesis H2 was also supported, indicating that AI adoption significantly enhances GHRM practices. This relationship highlights how digital technologies are transforming HR functions from administrative tasks into strategic enablers of sustainability (Alzoubi & Mishra, 2024; Soomro et al., 2024). AI-driven HR systems automate green recruitment processes, personalize sustainability training, and evaluate employee performance using environmental criteria.

From a theoretical standpoint, this finding demonstrates that AI not only contributes to operational efficiency but also enables strategic human capital development for sustainability. Within the RBV framework, GHRM becomes a valuable organizational capability, while AI provides the technological backbone for its optimization. Furthermore, according to SCT, AI-enabled systems influence employee learning and cognition by providing continuous feedback and interactive experiences that reinforce eco-conscious behaviors. The positive association between AI and GHRM can also be attributed to the growing recognition among SME managers that sustainability-focused HR policies improve reputation, employee engagement, and innovation. The integration of AI and HRM thus represents a dual pathway technological efficiency combined with human sustainability which ultimately drives firm performance and legitimacy in competitive markets.

H3: Green HRM → Environmental Sustainability

The study further confirms that GHRM positively influences environmental sustainability, validating H3. This result is consistent with prior empirical research (Chand et al., 2024; Shaikh et al., 2024; Sharma et al., 2022) showing that HR policies emphasizing environmental responsibility foster employee participation in green initiatives and resource conservation. Green recruitment ensures that new employees share the organization’s sustainability values, while green training enhances their knowledge of energy conservation, waste reduction, and eco-friendly production techniques.

From an SCT perspective, GHRM provides the social and behavioral context in which employees learn, model, and reinforce sustainable practices. It shapes the psychological contract between the organization and its workforce, signaling that environmental responsibility is a valued and rewarded behavior. From a strategic management perspective, these practices embed sustainability within the firm’s core competencies, ensuring that environmental performance is achieved not through compliance alone but through continuous learning and behavioral alignment.

In SMEs, where resource constraints often limit large-scale sustainability investments, GHRM emerges as a cost-effective mechanism to improve environmental outcomes through human-centered strategies. Hence, HRM becomes a bridge between corporate strategy and environmental performance, transforming employees from passive executors into active sustainability agents.

H4: AI Adoption → Green Employee Behavior

The results confirm that AI adoption has a positive and significant influence on GEB in SMEs, underscoring the role of digital technologies in shaping individual pro-environmental actions. This finding aligns with social cognitive theory (Marks & Bandura, 2002), which emphasizes that behavior is influenced by observation, self-efficacy, and reinforcement. Through intelligent systems such as AI-based training platforms, predictive analytics, and digital feedback tools employees observe, learn, and internalize environmentally responsible practices. AI thus acts as a behavioral enabler, facilitating awareness and motivation toward sustainability.

Empirically, the result resonates with Shafaei et al. (2020), who demonstrated that technology-supported environmental initiatives enhance employees’ eco-conscious behaviors through gamified learning and feedback mechanisms. Similarly, Aydin et al. (2024) observed that AI-driven HR systems improve engagement by tailoring training and recognition systems that reward sustainable conduct. Within the SME context, where limited resources often constrain formal sustainability programs, AI adoption substitutes for such gaps by providing real-time behavioral monitoring and individualized reinforcement. Therefore, consistent with prior evidence, this study validates that AI not only supports operational efficiency but also cultivates a sustainability-oriented workforce through cognitive and behavioral transformation.

H5: Green Employee Behavior → Environmental Sustainability

The study also establishes that GEB significantly contributes to ES in SMEs. This result is theoretically grounded in Social Cognitive Theory, which posits that sustained behavioral change at the individual level accumulates into organizational and systemic transformation (Marks & Bandura, 2002). Employees who conserve energy, minimize waste, and innovate eco-friendly practices directly translate organizational environmental intent into measurable sustainability outcomes.

The finding corroborates Shafaei et al. (2020), who identified GEB as the behavioral mechanism linking organizational strategies with ecological results, and Shaikh et al. (2024), who emphasized that proactive employee engagement enhances firms’ environmental and energy efficiency. Similarly, Sun et al. (2022) highlighted that organizations embedding environmental KPIs into performance evaluations experience stronger ecological results due to increased employee accountability. Within the SME context, where structural sustainability initiatives may be limited, employees’ voluntary actions serve as the micro-foundations of environmental performance. Collectively, these results affirm that GEB is not merely an outcome of HR or technological interventions but a central conduit through which environmental sustainability materializes in resource-constrained enterprises.

Mediation Effect

H6: Mediation of GHRM between AI Adoption and ES

The mediation analysis reveals that GHRM partially mediates the relationship between AI adoption and ES, indicating that AI’s impact on sustainability operates partly through its influence on HR systems and practices. This finding extends both the RBV and SCT by demonstrating that AI transforms HR processes into strategic sustainability capabilities. AI integration into HR functions, such as green recruitment, sustainability-focused training, and environmental performance evaluation enables organizations to align workforce management with ecological objectives (Aydin et al., 2024; Soomro et al., 2024). By automating these functions, AI helps firms institutionalize environmental criteria within hiring, appraisal, and learning systems, ensuring that sustainability becomes a recurring organizational norm rather than an episodic initiative. Moreover, consistent with Sulaiman et al. (2016), a digitally supported HR environment increases employees’ receptiveness to environmental initiatives and improves organizational learning.

In SMEs, where financial and technical resources are limited, AI-enabled GHRM offers a scalable mechanism for embedding environmental awareness and accountability. The mediation result, therefore, suggests that AI adoption indirectly enhances ES by strengthening HR systems that, in turn, influence employee commitment and sustainability-driven performance. This aligns with RBV’s assertion that organizational resources derive strategic value when embedded in firm-specific routines here, AI-driven GHRM routines that foster enduring environmental capabilities.

H7: Mediation of Green Employee Behavior

The mediating analysis confirmed that GEB significantly transmits the effect of AI adoption on environmental sustainability, supporting H7. This finding is theoretically meaningful because it identifies employee behavior as the mechanism through which technological and managerial systems translate into environmental outcomes. AI-enabled HR practices such as gamified sustainability training, digital performance tracking, and personalized feedback motivate employees to adopt pro-environmental behaviors like energy conservation and waste minimization (Reddy et al., 2024; Shafaei et al., 2020).

Under the SCT lens, this mediation validates the cognitive–behavioral pathway: AI tools influence individuals’ awareness, efficacy beliefs, and behavioral intentions, which in turn shape their actual environmental actions. Employees who perceive their organization as technologically advanced and sustainability-driven are more likely to engage voluntarily in green behaviors.

Moreover, the mediation result emphasizes that technology alone is not transformative its impact depends on the human capacity to internalize and operationalize sustainability goals. This insight enriches existing literature by showing that AI-driven sustainability is a social process, mediated by the workforce’s behavioral engagement. It suggests that future AI policies in SMEs should focus equally on technological capability building and behavioral motivation systems.

Moderation Effect

H8: Moderation of Organizational Culture

The results supported H7, revealing that OC significantly moderates the relationship between GHRM and ES. Specifically, a strong pro-environmental culture amplifies the effectiveness of GHRM practices, whereas a weak culture diminishes it. This finding resonates with Sheikh et al. (2024), Wang (2019), who demonstrated that organizations with deeply embedded sustainability values achieve superior environmental outcomes.

Theoretically, this aligns with OCT, which posits that shared beliefs and values shape collective behavior and performance. In SMEs, where informal relationships and leadership influence are strong, culture acts as the “social glue” binding employees to environmental goals. When leaders actively promote sustainability and reward green behaviors, employees perceive environmental performance as a shared responsibility, not an imposed directive.

This moderating effect underscores that GHRM cannot function effectively in a cultural vacuum. While policies and AI-enabled systems can institutionalize sustainability structures, culture determines the depth and durability of these practices. In this sense, OC is the contextual engine that determines whether sustainability remains a temporary initiative or evolves into an enduring organizational identity.

Integrated Theoretical Insights

Collectively, the findings of this study advance the integration of three major theoretical perspectives the RBV, SCT, and OCT to provide a holistic understanding of environmental sustainability in SMEs. From the RBV perspective, AI adoption and GHRM emerge as valuable, rare, and inimitable organizational resources that enable firms to achieve sustainability-driven competitive advantage. SCT further explains the behavioral mechanisms, demonstrating that GEB serve as the conduit linking technological and managerial practices with environmental outcomes. Meanwhile, OCT contextualizes these relationships by showing how shared values, norms, and cultural beliefs within organizations amplify the effectiveness of sustainability initiatives. Integrating these three perspectives offers a multi-level explanation of how digital transformation, human resource systems, individual behavior, and organizational culture collectively drive ecological resilience in SMEs. This synthesis contributes to theory by highlighting that sustainable performance results not merely from technological innovation, but from the alignment of technological capabilities, human engagement, and cultural support.

Practical Implications

This study demonstrates the practical potential of AI technology in promoting environmental sustainability among SMEs through process redesign, predictive analytics, and resource optimization. The integration of AI technologies allows SMEs to overcome key sustainability challenges, particularly those related to resource constraints and environmental impact. Moreover, the findings suggest that AI can be adopted gradually through scalable, low-cost technologies suitable for smaller enterprises, allowing firms to digitize operations without excessive financial burden.

In addition, GHRM enhanced by AI capabilities, plays a significant role in driving environmental sustainability. Through green recruitment, company-specific sustainability training and performance appraisals focused on environmentally driven objectives, SMEs can better integrate their workforce to achieve sustainability goals. These initiatives encourage employees to practice eco-friendly behaviors and foster a positive environment for sustainability.

Furthermore, leadership and organizational culture are crucial in ensuring that these initiatives succeed. Leaders must cultivate a sustainability-oriented culture, communicate clear environmental values, and motivate employees to engage actively with AI-powered environmental initiatives. A supportive culture ensures that sustainability becomes part of the organizational identity rather than a short-term project. Policymakers should also play a facilitating role by providing financial incentives, digital-skills programs, and sustainability-oriented training for employees to enhance AI adoption. Public–private partnerships and government subsidies can further reduce the technology-adoption gap for SMEs. These interventions will help firms implement AI-driven sustainability initiatives, ultimately contributing to a more resource-efficient and environmentally responsible manufacturing sector.

Theoretical Implications

This study makes several theoretical contributions to the field of AI, GHRM and sustainability. This research deepens SCT by showing how AI-enabled HR practices shape employee behaviors to achieve sustainability outcomes highlighting the importance of GEB in translating organizational sustainability strategies into measurable environmental improvements.

The research also extends the RBV by categorizing AI and GHRM as valuable organizational resources that drive sustainability values within corporate culture to maximize AI`s impact on environmental goals. Moreover, it builds on OCT by providing empirical evidence that organizational culture moderates the AI–GHRM–sustainability relationship, reinforcing the importance of embedding sustainability values within corporate culture.

Additionally, this research offers insights into how emerging markets can benefit from the synergistic relationship between AI, GHRM, and sustainability. These findings address a gap in the literature by demonstrating how digital transformation can drive sustainability in SMEs, particularly in developing economies. (Kamble et al., 2021; Soomro et al., 2024). Finally, this study proposes a comprehensive framework integrating AI, GHRM, GEB, OC, and ES, establishing a theoretical foundation for future sustainability research and clarifying how the alignment of technology, human capital, and culture jointly drives environmental performance.

Limitations and Future Research

Despite its significant contributions, this study has several limitations that provide opportunities for future research. First, this research focuses solely on SMEs in the manufacturing sector of Pakistan, limiting the generalizability of the findings. Future research should expand to other sectors (e.g., healthcare, education, services) to explore how AI and GHRM affect sustainability in diverse industries.

Second, the study employs a cross-sectional research design, which captures relationships at a single point in time. Future research should conduct longitudinal studies to assess dynamic changes in the interaction between AI, GHRM, and sustainability.

Third, this study identifies AI adoption and GHRM as key antecedents of sustainability, with GEB as the mediator and OC as the moderator. Future research could incorporate additional constructs, such as leadership, corporate governance, and employee engagement, to enhance understanding of sustainability outcomes. Future studies should examine other digital innovations (e.g., blockchain, digital twins, and machine learning) to determine how they complement AI-driven sustainability initiatives.

Lastly, this study acknowledges but does not empirically test the various ethical issues and dilemmas related to AI adoption, such as algorithmic bias, data privacy, and transparency in AI-driven decision-making. Future research should address these ethical dimensions to provide a more comprehensive understanding of AI’s role in sustainable development and responsible innovation.

Conclusion

This study investigated the relationships among AI, GHRM, GEB, OC, and ES within the SME manufacturing sector. By employing a PLS-SEM approach, the findings provide valuable insights into how SMEs can leverage AI and GHRM to drive sustainability initiatives effectively in resource-constrained environments.

The results emphasize that AI acts as a transformative tool, enabling SMEs to optimize resource management, enhance operational efficiency, and achieve sustainability goals. GHRM further strengthens this impact by embedding sustainability principles into HR practices such as green recruitment, tailored training, and eco-focused performance evaluations. Together, these constructs significantly influence GEB, which emerges as a critical mediator, ensuring that organizational strategies translate into tangible sustainability outcomes. Additionally, the study highlights the moderating role of OC, demonstrating that a strong pro-environmental culture amplifies the effectiveness of AI and GHRM in promoting ES. The study advances theoretical frameworks such as Social Cognitive Theory, Resource-Based View, and Organizational Culture Theory by integrating novel constructs like GEB and OC into the sustainability discourse. It also addresses gaps in existing literature by exploring the interplay between AI, GHRM, and sustainability in emerging markets, providing a modernized perspective on how SMEs can align technological innovation with environmental objectives.

From a practical standpoint, this study offers actionable strategies for SME leaders to adopt AI-driven technologies, foster pro-environmental organizational cultures, and engage employees in sustainability initiatives. The research also provides a roadmap for policymakers to design targeted interventions that promote AI adoption and sustainability practices, enabling SMEs to overcome barriers such as resource limitations and operational inefficiencies.

In conclusion, this study bridges the gap between technological innovation and environmental sustainability by demonstrating the synergistic effects of AI, GHRM, and OC on ES. By addressing critical gaps in the literature and offering practical insights, it provides a comprehensive framework for academics, practitioners, and policymakers seeking to enhance sustainability outcomes in SMEs. Future research can build on these findings by Exploring AI-driven sustainability strategies across diverse industries (e.g., healthcare, agriculture, and services), expanding cross-regional studies to understand the influence of cultural and economic differences on AI adoption in sustainability and investigating emerging AI innovations such as blockchain and digital twins to assess their potential in sustainability-driven business transformation. By integrating AI, HRM, and sustainability practices, SMEs can position themselves as leaders in sustainable business models, ensuring long-term environmental and economic resilience.

Footnotes

Appendix 1: Questionnaire

ORCID iDs

Sonia Najam Shaikh

Jan Muhammad Sohu

Fatima Zahra Kherazi

Ethical Considerations

Ethical approval and informed consent were not required for this study as per the institutional and national guidelines, because it involved anonymous voluntary survey responses without any intervention or personal data collection. All participants were informed of the purpose and confidentiality of the study.

Authors Contributions

Conceptualization: Sonia N. S., Jan M. S., and Fatima Z.K.; Methodology: Jan M. S. and Sonia N. S.; Validation: Jan M. S., Sonia N. S., and S.S; Formal Analysis: Jan M. S. and FZK.; Resources: Jan M. S. and Fatima Z. K.; Data Curation: Jan M. S. and Suman N. S.; Writing Original Draft Preparation: Jan M. S., Sanam S., and Fatima Z. K.; Writing Review and Editing: Sonia N. S. and Sanam S.; Project Administration: Suman N. S. and Sonia N. S.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

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

Data Availability Statement

The data will be available from the corresponding author upon reasonable request.

References

Ababneh

O. M. A.

(2021). How do green HRM practices affect employees’ green behaviors? The role of employee engagement and personality attributes. Journal of Environmental Planning and Management, 64(7), 1204–1226. https://doi.org/10.1080/09640568.2020.1814708

Akter

Michael

Uddin

M. R.

McCarthy

Rahman

(2022). Transforming business using digital innovations: The application of AI, blockchain, cloud and data analytics. Annals of Operations Research, 308(1–2), 7–39. https://doi.org/10.1007/S10479-020-03620-W/METRICS

Aliyah

C. L.

Pangilinan

G. A.

Chakim

M. H. R.

Saputra

D. B.

(2023). Examining the impact of artificial intelligence and internet of things on smart tourism destinations: A comprehensive study. Aptisi Transactions on Technopreneurship (ATT), 5(2sp), 135–145. https://doi.org/10.34306/att.v5i2sp.332

Alzoubi

Mishra

(2024). Green artificial intelligence initiatives: Potentials and challenges. Journal of Cleaner Production, 468, Article 143090. https://doi.org/10.1016/j.jclepro.2024.143090

Alzyoud

(2021). The impact of green human resource management practices and knowledge sharing on sustainable performance: a conceptual framework. International Journal of Education Humanities and Social Science, 4(2), 115–132.

Aydin

Karaarslan

Narin

N. G.

(2024). Artificial intelligence, VR, AR and metaverse technologies for human resources management. arXiv arXiv:2406.15383. https://doi.org/10.48550/ARXIV.2406.15383

Carrasco-Carvajal

García-Pérez-de-Lema

Castillo-Vergara

(2023). Impact of innovation strategy, absorptive capacity, and open innovation on SME performance: A Chilean case study. Journal of Open Innovation: Technology, Market, and Complexity, 9(2), Article 100065. https://doi.org/10.1016/J.JOITMC.2023.100065

Chand

Nijjer

Jandwani

Narula

G. S.

(2024). A novel funnel and ontological mechanism for sustainable green human resource management (GHRM). International Journal of Information Technology (Singapore), 16(1), 369–374.

Dolce

Vinzi

V. E.

Lauro

(2017). Predictive path modeling through PLS and other component-based approaches: Methodological issues and performance evaluation. In Latan

Hair

J. F.

Noonan

(Eds.), Partial least squares path modeling: Basic concepts, methodological issues and applications (pp. 153–72). Springer International Publishing.

10.

Gao

(2023). R-Squared (R²) – How much variation is explained? Research Methods in Medicine & Health Sciences, 5(4), 104–109. https://doi.org/10.1177/26320843231186398

11.

Garg

Srivastav

Gupta

(2018). Application of artificial intelligence for sustaining green human resource management [Conference session]. 2018 International Conference on Automation and Computational Engineering (ICACE), Greater Noida, India, pp. 113–116.

12.

Hair

J. F.

Gabriel

M. L. D. S.

da Silva

Junior

S. B.

(2019). Development and validation of attitudes measurement scales: Fundamental and practical aspects. RAUSP Management Journal, 54(4), 490–507. https://doi.org/10.1108/RAUSP-05-2019-0098

13.

Hair

J. F.

Tomas Hult

Ringle

Sarstedt

(2017). A primer on partial least squares structural equation modeling (PLS-SEM) (p. 374). SAGE Publications.

14.

Hongyun

Sohu

J. M.

Khan

A. U.

Junejo

Shaikh

S. N.

Akhtar

Bilal

(2023). Navigating the digital landscape: Examining the interdependencies of digital transformation and big data in driving SMEs’ innovation performance. Kybernetes, 52(12), 1–29. https://doi.org/10.1108/K-07-2023-1183

15.

Kamble

S. S.

Gunasekaran

Kumar

Belhadi

Foropon

(2021). A machine learning based approach for predicting blockchain adoption in supply chain. Technological Forecasting and Social Change, 163, Article 120465. https://doi.org/10.1016/j.techfore.2020.120465

16.

Kim

T. T.

Kim

W. G.

Majeed

Haldorai

(2023). Does green human resource management lead to a green competitive advantage? A sequential mediation model with three mediators. International Journal of Hospitality Management, 111, Article 103486. https://doi.org/10.1016/j.ijhm.2023.103486

17.

Kock

(2015). Common method bias in PLS-SEM. International Journal of E-Collaboration, 11(4), 1–10. https://doi.org/10.4018/IJEC.2015100101

18.

Lee

Suh

Roy

Baucus

(2019). Emerging technology and business model innovation: The case of artificial intelligence. Journal of Open Innovation: Technology, Market, and Complexity, 5(3), Article 44. https://doi.org/10.3390/JOITMC5030044

19.

Manika

Wells

Gregory-Smith

Gentry

(2013). The impact of individual attitudinal and organisational variables on workplace environmentally friendly behaviours. Journal of Business Ethics, 126, 663–684. https://doi.org/10.1007/s10551-013-1978-6

20.

Marks

Bandura

(2002). Social foundations of thought and action (pp. 94–106). SAGE Publications Ltd.

21.

Matloob

Limón

M. L. S

Montemayor

H. M. V.

Raza

Rodriguez

J. C. C.

(2023). Does strategic change enhance the relationship between firms’ resources and SMEs performance in Pakistan? Sustainability, 15(3), Article 1808. https://doi.org/10.3390/su15031808

22.

Qasim

Qureshi

M. A.

Shaikh

S. N.

(2025). Green management and sustainability in SMEs: A pathway to a greener economy. In Qureshi

M. A.

Qureshi

J. A.

Tunio

M. N.

(Eds.), Examining green human resources management and nascent entrepreneurship (pp. 115–144). IGI Global Scientific Publishing.

23.

Reddy

M. R. S.

Deepthi

Bhattaru

Srilakshmi

Singh

(2024). Harmony in HR: Exploring the synergy of artificial intelligence and green practices for sustainable workplaces. MATEC Web of Conferences, 392, Article 01039. https://doi.org/10.1051/matecconf/202439201039

24.

Schwäke

Peters

Kanbach

Kraus

Jones

(2024). The new normal: The status quo of AI adoption in SMEs. Journal of Small Business Management, 63(3), 1297–1331. https://doi.org/10.1080/00472778.2024.2379999.

25.

Shafaei

Nejati

Yusoff

Y. M.

(2020). Green human resource management: A two-study investigation of antecedents and outcomes. International Journal of Manpower, 41(7), 1041–1060. https://doi.org/10.1108/IJM-08-2019-0406

26.

Shaikh

S. N.

Zhen

Sohu

J. M.

Soomro

Akhtar

Kherazi

F. Z.

Najam

(2024). How green HRM practices foster green competitive advantage through big data analytics capability and are amplified by managerial environmental concern. Kybernetes. Advance online publication. https://doi.org/10.1108/K-06-2024-1433

27.

Sharma

Sakhuja

Nijjer

(2022). Recent trends of green human resource management: Text mining and network analysis. Environmental Science and Pollution Research, 29(56), 84916–84935. https://doi.org/10.1007/s11356-022-21471-9

28.

Sheikh

A. A.

Shan

Hassan

N. M.

Khan

S. N.

AbdAlatti

(2024). Impact of green human resource management practices on hotels environmental performance: A mediation and moderation analysis. Sustainable Futures, 8, Article 100409. https://doi.org/10.1016/j.sftr.2024.100409

29.

Shmueli

Ray

Estrada

J. V.

Chatla

(2016). The elephant in the room: Predictive performance of PLS models. Journal of Business Research, 69(10), 4552–4564. https://doi.org/10.1016/j.jbusres.2016.03.049

30.

Sipola

Saunila

Ukko

(2023). Adopting artificial intelligence in sustainable business. Journal of Cleaner Production, 426, Article 139197. https://doi.org/10.1016/j.jclepro.2023.139197

31.

Sohu

J. M.

Mirani

M. A.

Shaikh

S. N.

Memon

Bhanbhro

Kherazi

F. Z.

(2025). Integrating green human resource management and big data analytics capability for sustainable environmental performance in SMEs. In Qureshi

M. A.

Qureshi

J. A.

Tunio

M. N.

(Eds.), Examining green human resources management and nascent entrepreneurship (pp. 145–172). IGI Global Scientific Publishing.

32.

Soomro

Fan

Sohu

J. M.

Soomro

Shaikh

S. N.

(2024). AI adoption: A bridge or a barrier? The moderating role of organizational support in the path toward employee well-being. Kybernetes. Advance online publication. https://doi.org/10.1108/K-07-2024-1889

33.

Sulaiman

Naqshbandi

M. M.

Dezdar

(2016). Impact of adoption of green IT practices on organizational performance. Quality & Quantity, 50, 1929–1948. https://doi.org/10.1007/s11135-015-0244-7

34.

Sun

Samad

Rehman

S. U.

Usman

(2022). Clean and green: The relevance of hotels’ website quality and environmental management initiatives for green customer loyalty. British Food Journal, 124(12), 4266–4285. https://doi.org/10.1108/BFJ-09-2021-1002

35.

Tenenhaus

Vinzi

V. E.

Chatelin

Y.-M.

Lauro

(2005). PLS path modeling. Computational Statistics & Data Analysis, 48(1), 159–205. https://doi.org/10.1016/j.csda.2004.03.005

36.

Wang

C.-H.

(2019). How organizational green culture influences green performance and competitive advantage: The mediating role of green innovation. Journal of Manufacturing Technology Management, 30(4), 666–683. https://doi.org/10.1108/JMTM-09-2018-0314

37.

Waqas

Yahya

Ahmed

Rasool

Hongbo

(2021). Unlocking employee’s green behavior in fertilizer industry: The role of green HRM practices and psychological ownership. International Food and Agribusiness Management Review, 24, 1–18. https://doi.org/10.22434/IFAMR2020.0109.

Artificial Intelligence Adoption,Green HRM,and Employee Behavior: Driving Environmental Sustainability in SMEs with Organizational Culture as a Moderator

Abstract

Keywords

Introduction

Literature Review

Artificial Intelligence Adoption and Environmental Sustainability

AI Adoption and Green HRM

GHRM and Environmental Sustainability

AI Adaption Positively Contributes GEB

GEB Positively Contributes to ES

Mediating Role of GHRM and GEB

Moderation of Organizational Culture

Methods

Sample and Data Collection

Variables and Measurement

Pilot Test and Questionnaire Modifications

Measurement Scale

Data Analysis

Results

Testing for Convergent and Discriminant Validity

Discriminant Validity

Inner Model VIF and f-Square

Model Fit, Predictive Validity, and R-Square

Hypothesis Testing and Mediation Effects

Structural Equation Model (SEM)

Mediation and Moderation Effects

Discussion

Direct Effects

H1: AI Adoption → Environmental Sustainability

H2: AI Adoption → Green Human Resource Management

H3: Green HRM → Environmental Sustainability

H4: AI Adoption → Green Employee Behavior

H5: Green Employee Behavior → Environmental Sustainability

Mediation Effect

H6: Mediation of GHRM between AI Adoption and ES

H7: Mediation of Green Employee Behavior

Moderation Effect

H8: Moderation of Organizational Culture

Integrated Theoretical Insights

Practical Implications

Theoretical Implications

Limitations and Future Research

Conclusion

Footnotes

Appendix 1: Questionnaire

ORCID iDs

Ethical Considerations

Authors Contributions

Funding

Declaration of Conflicting Interests

Data Availability Statement

References