Latest / 1 More Paper / Achieving green competitive advantage through generative AI: the mediating roles of organisational creativity and green innovation ambidexterity in manufacturing — Ruangkanjanases et al. (2025)
Transcript
- You’re listening to “Achieving green competitive advantage through generative AI: the mediating roles of organisational creativity and green innovation ambidexterity in manufacturing,” by A. Ruangkanjanases and colleagues. Published in 2025. International Journal of Logistics Research and Applications A Leading Journal of Supply Chain Management ISSN: 1367-5567 (Print) 1469-848X (Online) Journal homepage: the linked source Achieving green competitive advantage through generative AI: the mediating roles of organisational creativity and green innovation ambidexterity in manufacturing Athapol Ruangkanjanases, Shih-Chih Chen, Ornlatcha Sivarak & Asif Khan To cite this article: Athapol Ruangkanjanases, Shih-Chih Chen, Ornlatcha Sivarak & Asif Khan (19 Oct 2025): Achieving green competitive advantage through generative AI: the mediating roles of organisational creativity and green innovation ambidexterity in manufacturing, International Journal of Logistics Research and Applications, DOI: 10.1080/13675567.2025.2573664 Achieving green competitive advantage through generative AI: the mediating roles of organisational creativity and green innovation ambidexterity in manufacturing b,c, Ornlatcha Sivarakd and Asif Khane Athapol Ruangkanjanasesa, Shih-Chih Chen aChulalongkorn Business School, Chulalongkorn University, Bangkok, Thailand; bDepartment of Information Management, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan; cInternational Master of Business Administration program, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan; dMahidol University International College, Mahidol University, Phutthamonthon, Nakhon Pathom, Thailand; eCollege of Business, Southern Taiwan University of Science and Technology, Tainan, Taiwan ABSTRACT. This study examines the implementation pathway of Generative AI Capabilities (GAIC) in manufacturing organisations, focusing on the roles of organisational creativity and green innovation ambidexterity in achieving a green competitive advantage. Drawing on the Technology-Organisation-Environment (TOE) framework, this research develops and empirically tests an integrated model using data collected from 297 senior and middle-level managers of manufacturing firms in Taiwan. The results indicate that both technological and organisational contexts have a significant influence on GAIC implementation, whereas the environmental context shows no significant impact. Furthermore, GAIC demonstrates significant positive effects on both organisational creativity and green innovation ambidexterity, which in turn enhances green competitive advantage. This study addresses critical research gaps by making several contributions by extending the TOE framework to encompass GAIC, thereby advancing the understanding of human-AI collaborative creativity. The findings provide novel insights into how manufacturing organisations can strategically implement GAIC to achieve a green competitive advantage. 1. Introduction. ARTICLE HISTORY Received 29 June 2025 Accepted 7 October 2025 Generative AI capabilities; technological context; organisational context; environmental context; green innovation ambidexterity; green competitive advantage In the context of Industry 5.0, artificial intelligence (AI) applications represent a cornerstone technological advancement. Manufacturing systems that incorporate cognitive simulation and adaptive learning yield significant operational benefits. These operational advantages extend innovative capabilities across industrial operations, market development, supply acquisition, distribution networks, and customer experience enhancement. The manufacturing sector’s AI positioning can be categorised into two principal areas. The initial category encompasses decision assistance frameworks that support management and operational staff without complete automation, providing analytical support for process refinement and system modeling while retaining essential human direction. The subsequent category comprises self-operating AI platforms that execute tasks without human intervention, enabling advanced production automation. Generative Artificial Intelligence capabilities (GAIC) stand out as a transformative advancement in recent technological developments. This innovation demonstrates exceptional proficiency in creating original textual, visual, and video-based materials. The structural framework of GAIC utilises sophisticated learning protocols to build and strengthen its knowledge repository. Contemporary platforms, particularly ChatGPT, demonstrate GAIC’s inherent capacity for self-directed learning and knowledge expansion, offering substantial value for organisational decision processes and management functions. Scholarly investigations have established GAIC’s vital role in advancing data analytics, strengthening supply network cohesion, and fostering unprecedented creative potential through original content development. In the current manufacturing environment, achieving green competitive advantage (GCA) has become both vital and challenging. GCA, grounded in the resource-based view (RBV), refers to an enterprise’s ability to sustain superior performance through strategically valuable, distinctive, non-replicable, and irreplaceable resources, while addressing ecological and societal imperatives. Manufacturing firms face increasing demands from environmental legislation, growing stakeholder expectations for sustainability initiatives, and intensified marketplace competition for environmentally responsible products. Balancing ecological outcomes with business performance presents substantial operational challenges, particularly in resource-intensive production sectors. Traditional pathways to GCA frequently encounter barriers in processing environmental data, developing innovations, and pursuing simultaneous incremental and transformative sustainability improvements. GAIC offers a promising avenue to address these challenges through its computational sophistication and inventive potential. Through advanced algorithmic systems and adaptive learning mechanisms, GAIC enables the analysis of comprehensive environmental information, improves resource deployment, and develops innovative approaches for sustainable manufacturing operations. These capabilities help organisations bridge the traditional gap between environmental and economic objectives by strengthening decision processes, improving operational systems, and advancing sustainable production methods. However, a significant knowledge gap remains in understanding the systematic implementation of GAIC for achieving GCA. The Technology-Organisation-Environment (TOE) framework establishes a systematic foundation for examining technological adoption within organisational contexts (wael AL-khatib 2023). Although scholars have widely utilised this framework to investigate diverse technological developments, its analytical application to GAIC within sustainable production environments remains largely unexamined. Furthermore, the foundational TOE structure, when applied independently, provides insufficient explanatory power regarding the transformational pathways through which GAIC cultivates GCA. Scholarly researchers and organisational practitioners are keenly interested in understanding adoption pathways that enhance operational performance and workforce effectiveness. Within the foundational theoretical frameworks examining organisational technology integration, the Technology-Organisation-Environment (TOE) framework delineates three key dimensions: technological, organisational, and environmental. This conceptual structure has received substantial scholarly recognition for thoroughly examining the internal and external elements that shape organisational technology decisions. The environmental context (ENC) and technological context (TEC) represent external determinants that influence an organisation’s capacity for technology integration, while the organisational context (ORC) encompasses internal elements crucial to the success of implementation. This investigation employs the TOE framework, acknowledging that the dynamic interactions between TEC, ORC, and ENC elements significantly contribute to meaningful adoption outcomes. These dimensional components may also amplify the intricacies and obstacles organisations encounter during technology integration processes. Current scholarly work acknowledges the TOE framework’s value as an analytical foundation for assessing organisational adoption of advanced AI systems and GAIC technologies (wael AL-khatib 2023). Accordingly, this study examines the relationships between TOE factors and the implementation of GAIC. This investigation addresses the theoretical gap by positioning organisational creativity (OCRE) and green innovation ambidexterity (GIA) as essential intermediary mechanisms in the analytical framework. These organisational competencies hold particular significance as they embody the intellectual and developmental pathways that facilitate the conversion of GAIC into enduring GCA. By incorporating these mediating constructs, this research expands the explanatory scope of the TOE framework, illuminating the transitional sequence from technological proficiency to ecological performance outcomes. OCRE, an enterprise’s capacity to develop novel and valuable concepts for products, services, and operational processes, has become essential within manufacturing sectors that confront technological evolution and sustainability requirements. Within production environments, OCRE manifests through innovative design concepts, process enhancements, and inventive responses to manufacturing complexities. The advent of GAIC technologies presents distinctive opportunities for enhancing OCRE within manufacturing enterprises by enabling the evaluation of extensive production data repositories, proposing innovative design variations, and identifying patterns that may elude human observation, potentially catalyzing creative solutions in industrial settings. This technological framework’s capacity to generate multiple solution alternatives while accounting for manufacturing parameters may substantially expand organisational creative possibilities. Nevertheless, despite its strategic significance, the association between these GAIC and OCRE within production environments remains insufficiently examined. Understanding these dynamics becomes particularly relevant as manufacturing enterprises increasingly utilise computational innovations to address complex challenges in sustainable production, supply chain network optimisation, and product customisation. Furthermore, while existing literature has investigated conventional AI applications in manufacturing, the distinctive attributes of GAIC, including their capacity to create, synthesise, and transform concepts suggest unique mechanisms for influencing OCRE compared to traditional approaches. Consequently, examining the influence of these GAIC on OCRE within manufacturing contexts proves essential for theoretical advancement and practical implementation. This investigation seeks to explore the relationship between GAIC and OCRE. Empirical investigations within AI and innovation domains substantiate the pivotal role of GAIC applications, specifically ChatGPT and Chatbots, in identifying innovative operational approaches. These technological implementations advance organisational capacity for digital supply chain enhancement through GIA. GIA theory delineates two distinct innovation categories: green exploratory innovation (GEP), encompassing fundamental environmental advancements that require extensive research investment and sophisticated organisational resources, thereby enabling substantial enhancements to sustainable supply chain capabilities. Conversely, green exploitative innovation (GET) yields strategic organisational benefits through leveraging established competencies, fostering resource-efficient and cost-effective environmental innovations (wael AL-khatib 2023). Consequently, improving operational efficiency. The introduction of GAIC establishes a pivotal opportunity for advancing both dimensions of GIA within manufacturing enterprises. Through sophisticated computational processes, GAIC facilitates GEP by conceptualising innovative sustainable designs, evaluating ecological material configurations, and uncovering novel environmental production methodologies (wael AL-khatib 2023). Concurrently, it strengthens GET by refining current sustainable practices, enhancing resource utilisation, and advancing environmental management frameworks. The interrelation between GAIC and GIA is significant in manufacturing contexts, where ecological imperatives converge with operational requirements. This investigation, therefore, examines the relationship between GAIC and GIA. Within the manufacturing sector, attaining GCA requires a balance between economic outcomes, environmental stewardship, and social responsibility. OCRE advances GCA by enabling manufacturers to conceptualise distinctive, environmentally conscious solutions that resist competitive replication. Through innovative processes, manufacturing enterprises develop novel approaches to sustainable production methodologies, waste minimisation protocols, and resource optimisation frameworks, thereby establishing distinctive market positions that align with environmental requirements. Concurrently, GIA strengthens GCA through complementary environmental innovation pathways. The synergy between these capabilities holds particular significance within manufacturing contexts, where environmental imperatives, resource limitations, and competitive forces converge. Furthermore, as sustainability emerges as a crucial differentiator in global markets, the effective positioning of OCRE and GIA may determine which manufacturers establish enduring competitive positions within an environmentally conscious commercial landscape. Consequently, this investigation examines the influence of OCRE and GIA on GCA. Although research interest in GAIC continues to expand, fundamental theoretical uncertainties remain regarding the transformational pathways that convert these capabilities into GCA. The TOE framework, despite its widespread application in technological implementation research, has yet to be thoroughly examined in the context of GAIC-enabled sustainability outcomes. Contemporary scholarship has primarily concentrated on examining direct relationships between technological capacities and enterprise performance, neglecting critical intermediate mechanisms. This research addresses these conceptual limitations by proposing an innovative dual-mediation structure that encompasses OCRE and GIA (wael AL-khatib 2023). Existing research examining AI implementation and environmental performance has largely emphasised qualitative investigations and theoretical frameworks, revealing a significant empirical deficit in validating the intricate relationships between GAIC and GCA. This investigation addresses this methodological shortcoming through systematic quantitative analysis, drawing upon data collected from manufacturing organisations in Taiwan. The analytical framework employs partial least squares structural equation modeling (PLS-SEM) to simultaneously test the direct and indirect effects of GAIC implementation through OCRE and GIA on GCA. The investigation employs the Anderson and Gerbing approach to assess the reliability of constructs and conduct an empirical examination of the established hypotheses. The scholarly contribution of this investigation covers the following research questions. First, it broadens the TOE framework’s analytical scope by investigating the influence of TEC, ORC, and ENC elements on the placement of GAIC within sustainable production contexts. Secondly, this study establishes novel theoretical connections of GAIC with OCRE and GIA within the domain of AI-driven sustainability. Third, it investigates the impacts of OCRE and GIA on GCA. Ultimately, this research presents an integrated theoretical framework that outlines the multifaceted progression from GAIC to GCA, thereby enriching both the literature on technology implementation and sustainable operations. The structure of this paper is organised as follows: Section 2 presents a literature review and hypothesis development, grounded in the TOE framework and literature on GAIC implementation. Section 3 outlines the methodology, describing the data collection from Taiwanese manufacturing firms. Section 4 presents the empirical findings on the relationships between TEC, ORC, and ENC with GAIC implementation and their subsequent effects on GCA. Section 5 discusses the theoretical contributions and managerial implications of the findings, while Section 6 provides the conclusions of this research. This study advances the understanding of how manufacturing organisations can effectively leverage GAIC to achieve GCA through the mediating roles of OCRE and GIA. 2. Literature review and hypothesis development. 2.1. TOE and generative AI capabilities. The TOE framework establishes a multifaceted analytical structure for examining the adoption and integration of GAIC within organisational settings. Since its introduction by Tornatzky and Fleischer, the TOE framework has exhibited consistent efficacy in elucidating how technical preparedness, institutional attributes, and contextual dynamics influence technological advancements. Within GAIC implementation, the TEC encompasses the advancement of algorithmic systems, computing architecture, and information processing competencies that constitute the foundational elements of generative AI platforms. The OC reflects internal determinants, including enterprise scale, technical proficiency, leadership endorsement, and institutional learning capacity that shape GAIC deployment. The ENC captures external influences, comprising regulatory mandates, competitive forces, and stakeholder sustainability expectations that propel GAIC adoption. This three-dimensional framework is particularly relevant for GAIC implementation as it acknowledges the intricate interactions among TEC, ORC, and ENC. The framework’s emphasis on contextual elements aligns precisely with GAIC’s distinctive requirements for unique technical infrastructure, institutional competencies, and environmental considerations essential for effective implementation. The TEC describes the internal and external technology features that shape adoption decisions. Research identifies relative advantage as a key component within TEC. Relative advantage measures how superior a new technology proves to be compared to current organisational systems and their associated benefits. Compatibility is another critical factor in technology acceptance, measuring how well new technological solutions align with a firm’s current operations. In the TEC framework, compatibility refers to whether the technology aligns with the organisation’s cultural elements and operational methods. Firms can modify their operational guidelines to support more substantial alignment between compatibility requirements and GAIC implementation. Empirical evidence from recent studies illuminates the profound influence of TEC factors on the successful implementation of GAIC. Organisations with advanced IT infrastructure and comprehensive technical resources demonstrate markedly enhanced capacity for GAIC placement and maintenance. Notably, technological readiness – characterised by sophisticated hardware configurations and software architectures – emerges as a decisive factor in determining the quality and reliability of generative AI applications. The maturity of existing technical frameworks proves instrumental in facilitating seamless GAIC integration and operational scalability. This relationship becomes particularly evident in environments that require intensive data processing, where technological sophistication directly influences the effectiveness of generative model development and refinement. Substantial empirical findings consistently validate that organisations possessing well-developed technological foundations achieve demonstrably higher outcomes in GAIC implementation, thereby establishing the TEC as a cornerstone of successful GAIC initiatives. When organisations recognise clear connections between GAIC and their existing procedures, they demonstrate a greater willingness to adopt this technology across various operational areas. These observations lead to the following hypothesis: H1. TEC significantly impacts GAIC. This research considers two ORC factors – organisational readiness and management support – influencing GAIC adoption. Management support refers to the extent to which organisational leaders endorse technological innovations. These leaders bridge the gap between individual and organisational technology acceptance, where the likelihood of adoption correlates with senior management’s innovative mindset. Research has documented the significant impact of management support on the success of innovation. The ORC framework also emphasises organisational readiness as a crucial component, measuring a firm’s preparedness and motivation to adopt new technologies. Such readiness manifests in an organisation’s capability to manage and fund new technological implementations. Scholars highlight organisational readiness as central to the integration of GAIC in business analytics. Empirical research has revealed compelling evidence that ORC plays a pivotal role in shaping the successful implementation of GAIC. Organisations distinguished by their advanced knowledge management frameworks and judiciously aligned resource allocation mechanisms consistently achieve higher rates of success in GAIC initiatives. The synergistic relationship between ORC and GAIC becomes particularly apparent when examining how organisational architecture and process alignment facilitate the seamless integration of generative AI solutions. Studies consistently demonstrate that institutions investing in methodical preparation and systematic capacity building realise markedly better outcomes in their GAIC placement efforts. The success of organisational implementation is strongly associated with organisational competencies and readiness factors, with change management expertise and strategic alignment emerging as crucial determinants. These insights extend the understanding of ORC’s influence beyond simple resource availability, highlighting its fundamental role in enhancing organisational preparation and capability. Evidence points to robust connections between organisational readiness and technology acceptance, confirming its role as a key determinant of GAIC adoption. Based on this foundation, the following hypothesis is proposed: H2. ORC significantly impacts GAIC. The ENC encompasses factors beyond an organisation’s control (wael AL-khatib 2023). These external elements respond to changing environmental dynamics when operating in the broader business environment. Government policies emerge as critical external forces that influence GAIC adoption decisions within the TOE framework. Such policies create conditions that support or inhibit organisations’ technology adoption efforts. Scholarly work underscores the government’s role in shaping technology adoption trajectories. Both governmental bodies and industry players facilitate GAIC adoption by offering training initiatives, technical support, customised recommendations, and various incentives. Organisations show higher GAIC implementation levels when supported by government regulations, policies, and legislation that promote technological advancement. Recent empirical studies have shed new light on how ENC shaped the evolution of GAIC. Competitive pressures emerge as a powerful promoter in determining both the rate of GAIC adoption and its successful implementation, particularly in fostering innovations that balance the exploitation of existing capabilities with exploration of new possibilities. The dual forces of market uncertainty and competitive dynamics serve as critical external drivers, fundamentally influencing how organisations approach GAIC development and placement while shaping their absorptive capacity and implementation approaches. The growing intensity of competition in digital markets has become increasingly influential in determining the path of GAIC development across industries. The broader environmental landscape creates dynamic pressures that guide organisations’ efforts to cultivate and enhance their GAIC, especially in leveraging these capabilities for competitive advantage and innovation. These findings highlight the sophisticated and multifaceted nature of the relationship between ENC and GAIC implementation. Studies on the implementation of GAIC reveal that governmental support mechanisms and incentive structures drive increased acceptance and adoption rates of GAIC. These findings lead to the following hypothesis: H3. ENC significantly impacts GAIC. 2.2. Generative AI capabilities and organisational creativity. Existing research examining AI’s business applications highlights its capacity to enhance OCRE and improve performance outcomes. Multiple sector-specific examples reveal quantifiable improvements in OCRE following GAIC adoption. Though limited in scope, these cases establish clear links between GAIC deployment and enhanced creative processes within organisations. Meta-analytic findings have recently emerged, strengthening the theoretical understanding and demonstrating a meaningful correlation between human-GAIC collaboration and enhanced OCRE performance across diverse organisational domains. Field research reveals how the systematic adoption of GAIC accelerates OCRE, promoting both incremental improvements and transformative innovations within organisational settings. This phenomenon is particularly evident in creative industries, where a compelling majority of organisations report substantial enhancements in their creative output following GAIC integration. Moreover, studies indicate that organisations cultivating environments that embrace experimentation witness tangible improvements in both their innovative capabilities and creative problem-solving approaches through the implementation of GAIC. By automating resource-intensive operations, GAIC technology frees up valuable human capital for creative pursuits, thereby broadening an organisation’s innovative capacity. The strategic implementation of GAIC enables leadership to uncover new insights by detecting patterns and correlations within complex datasets that were previously unattainable. Contemporary studies describe multiple scenarios in which GAIC-generated insights have sparked innovative organisational solutions. GAIC platforms demonstrate significant potential for advancing OCRE within organisational frameworks. Based on this analytical foundation, the following hypothesis is proposed: H4. GAIC significantly impacts OCRE. 2.3. Generative AI capabilities and green innovation ambidexterity. Organisational ambidexterity literature positions innovation as a critical organisational outcome. GEP encompasses intensive research and development initiatives, reflecting a strategic orientation towards breakthrough innovations that represent market novelty. This approach requires substantial capital allocation and carries a higher risk potential. Conversely, GET focuses on incremental product or service innovations, characterised by modest modifications and a more conservative innovation strategy that builds capabilities while maintaining lower risk levels and reduced research and development expenditure. Optimal equilibrium between GEP and GET activities yields enhanced organisational returns and sustainable competitive advantages. Studies demonstrate that AI capabilities enhance organisations’ capacity for innovation ambidexterity (wael AL-khatib 2023). The implementation of AI technology broadens methodological exploration, augments research and development capabilities, and creates novel innovations, reinforcing GEP activities. Moreover, GAIC contributes to self-learning systems, maximises resource efficiency, strengthens core competencies, facilitates product standardisation, improves operational procedures, and delivers incremental innovations, supporting GET initiatives. Contemporary empirical research provides compelling evidence for the transformative role of GAIC in advancing GIA. In their seminal work, Wang and Zhang (2025) establish the fundamental contribution of GAIC to sustainable business model innovation through enhanced dynamic capabilities. Building on this foundation, further research has documented significant improvements in environmental performance metrics, attributing these enhancements to GAIC’s resource composition mechanisms. Furthermore, additional investigation further enriched these findings, demonstrating that the integration of GAIC with ambidextrous leadership yielded an enhancement in overall sustainability performance metrics. Collectively, this body of evidence highlights GAIC’s pivotal role in promoting both the GET and GEP dimensions of GIA. Drawing from this theoretical grounding regarding GAIC’s constructive impact on GIA, the following hypothesis is proposed: H5. GAIC significantly impacts GIA. 2.4. Organisational creativity and green competitive advantage. GCA manifests through an organisation’s unique combination of attributes that shape its approach to environmental management and sustainable innovation practices. Leadership enhances OCRE by fostering employee engagement with corporate information and knowledge assets. Entities exhibiting robust OCRE maintain long-term viability, consequently reinforcing operational effectiveness and GCA. Moreover, exceptional environmental achievement metrics signify established GCA. Organisations demonstrate their commitment to OCRE by creating and offering environmentally responsible products and services. These initiatives exemplify workforce participation in environmentally focused innovation efforts. Staff members who display OCRE exhibit sustained dedication to environmental conservation while working towards GCA advancement. Empirical evidence suggests that heightened OCRE significantly shapes GCA formation. The presence of individual inventive potential within organisational structures substantively enhances sustained commercial viability and organisational GCA. Recent studies examining manufacturing firms reveal that OCRE contributes to GCA through specific pathways. Manufacturing firms with high levels of OCRE demonstrated superior environmental performance through novel approaches to green product design and eco-efficient production processes. In addition, according to Demir et al., OCRE enables manufacturers to develop unique environmental management processes that directly enhance their competitive position by reducing resource consumption and implementing innovative waste management solutions. The relationship between OCRE and GCA is particularly pronounced in resource-intensive manufacturing sectors, where creative solutions to environmental challenges directly translate into GCA through cost reduction and market differentiation. These findings establish a clear theoretical and empirical foundation for the relationship between organisational creativity and green competitive advantage in manufacturing contexts. Hence, the following hypothesis can be proposed. H5. OCRE significantly impacts GCA. 2.5. Green innovation ambidexterity and green competitive advantage. According to RBV, GCA develops through an organisation’s distinctive resources and internal competencies. The theory emphasises that GCA stems from resources exhibiting four essential attributes: rarity, value creation, resistance to imitation, and lack of substitutes. Research extends this framework by identifying how capabilities in pollution reduction, sustainability initiatives, and environmental problem-solving yield competitive benefits. GIA represents a key organisational capability that drives GCA development. Through GIA implementation, firms create market distinctions by enhancing product design and quality, thereby achieving GCA. The application of GIA yields operational efficiencies through improved material usage, reduced energy consumption, enhanced waste processing, and optimised resource allocation. These cost reductions strengthen the firm’s GCA position. Although research on the manufacturing sector specifically examining GIA and GCA relationships remains relatively nascent, adjacent research streams offer valuable perspectives. The existing literature documents positive relationships between environmental management practices, environmental capabilities, and proactive environmental approaches, as well as GCA. Recent scholarly work has deepened our understanding of the theoretical linkages between GIA and GCA. In a notable contribution, Purnomo reveals how organisations leveraging green innovation capabilities through ambidexterity achieve distinctive GCA market positions by excelling in both environmental innovation and operational efficiency. Building on this foundation, Li et al. establish a compelling positive correlation between GIA and green dynamic capabilities, while Asiaei et al. demonstrate how the strategic deployment of green assets through ambidextrous approaches fosters GCA. This growing body of evidence provides substantial support for the hypothesis that GIA has a significant impact on GCA through multiple validated pathways. This research positions GIA as a fundamental capability advancing GCA. From this theoretical base emerges the following hypothesis: H5. GIA significantly impacts GCA. The theoretical framework of the study is presented in Figure 1. 3. Methodology. 3.1. Data collection and sample. This study employed a multi-stage sampling approach to collect data from manufacturing industries in Taiwan. The data collection process was conducted between September and December 2024, utilising SurveyMonkey, a widely used professional online survey platform in Taiwan. This platform was chosen for its ability to reach a diverse pool of respondents and its built-in quality control mechanisms. The study employed SurveyMonkey’s enterprise platform for data collection, leveraging its sophisticated B2B panel access and methodological capabilities to ensure reliable sampling and data integrity. This methodological choice aligns with contemporary approaches in organisational research and addresses the inherent challenges of accessing decision-makers in the manufacturing sector. The platform’s multi-tiered verification protocols and response validation mechanisms enhanced data quality through systematic screening of respondents based on predetermined criteria, including organisational role, experience level, and involvement in implementation decisions. This approach proved particularly valuable given the study’s focus on GAIC and GCA, where respondent expertise significantly influences data validity. Additionally, the platform’s established presence in Taiwan’s business ecosystem facilitated access to a demographically appropriate sample of manufacturing executives, thereby supporting the study’s geographical and sectoral focus. This research employed a stratified sampling methodology. The sampling process involved several steps. Initially, this study identified potential respondents using SurveyMonkey’s B2B panel, which consists of verified professionals from manufacturing enterprises. The screening criteria included employees working in manufacturing industries, a minimum of 2 years of work experience, familiarity with their organisation’s technological initiatives, and involvement in decision-making processes. While SurveyMonkey was the primary data collection platform, this study implemented a multi-channel approach to enhance data quality and representativeness. Additionally, this research collaborated with LinkedIn to reach senior executives and decision-makers in the manufacturing sector. To ensure geographical diversity, the research worked with regional industrial associations in Taipei, Taichung, Tainan, and Kaohsiung to distribute surveys through their official communication channels. This multi-channel approach increased the response rate and helped mitigate potential platform-specific biases. To maintain consistency across different collection methods, this research standardised the survey format and implemented identical screening criteria and quality control measures across all channels. The responses collected through these additional channels (approximately 25% of the total valid responses) were compared with those from SurveyMonkey to ensure there were no significant differences in response patterns. To ensure representative coverage, the study stratified the sample by company size (small, medium, and large enterprises), manufacturing sector (traditional manufacturing and high-tech manufacturing), geographic region (Northern, Central, and Southern Taiwan), and organisational position (executives, senior management, and middle management). Quality control measures included implementing attention-check questions, time controls for survey completion (with an expected duration of 15–20 min), IP address verification to prevent duplicate responses, and reverse-coded items to verify response consistency. The survey was distributed to 500 qualified participants to account for potential invalid responses and ensure the target sample size of 297 responses was achieved. Participants received a detailed introduction to the study’s purpose and assurances of confidentiality. 3.2. Sample characteristics. From the 500 distributed surveys, we received 338 responses (67.6% response rate). After removing incomplete responses and those failing attention checks, 297 valid responses remained. The final sample demonstrated diverse characteristics across organisational positions, industry distribution, company size, and geographic location. The sample comprised organisational positions, including senior executives (15%), middle management (45%), technical specialists (25%), and operational managers (15%). The industry distribution showed representation from traditional manufacturing (45%), high-tech manufacturing (35%), and smart manufacturing (20%). Regarding company size, large enterprises with over 1000 employees constituted 35% of the sample, medium enterprises with 300–1000 employees represented 40%, and small enterprises with fewer than 300 employees accounted for 25%. The geographic distribution covered Northern Taiwan (45%), Central Taiwan (35%), and Southern Taiwan (20%). Additionally, we conducted Harman’s single-factor test to check for common method bias. The results showed that no single factor accounted for more than 30% of the variance, indicating that common method bias was not a significant concern. 3.3. Data collection instrument. This research used a 7-point Likert scale. The research items to measure GAIC were adapted from wael Al-khatib & Khattab’s study. Furthermore, TOE constructs were measured by the items suggested by Chen et al.’s research. OCRE was measured by Mikalef & Gupta’s research items. GIA was measured by the items suggested by Chen, Khan, et al.. Lastly, the items to measure GCA were adapted from Singh et al.’s research. 4. Data analysis. The data were examined through a two-phase analytical sequence using partial least squares (PLS) methodology. The initial phase concentrated on verifying construct validity and reliability parameters, followed by a systematic assessment of path coefficients and inter-construct causal relationships. The adoption of PLS methodology was predicated on its established efficacy in preserving theoretical relationship structures while effectively managing complex research frameworks. This analytical framework demonstrates particular utility in scenarios characterised by non-normal distributions, incorporating specialised measurement protocols to address randomisation effects in the data. The research employed sequential analytical procedures,, utilising PLS-SEM’s inherent capability to evaluate multifaceted model configurations. 4.1. Convergent validity. Multiple measurement criteria were employed to assess convergent validity. Factor loadings and Cronbach’s alpha served as measures of internal consistency, complemented by RhoA and composite reliability (CR) as additional indicators of reliability. The RhoA measurement evaluates reliability through weight-based calculations rather than loading considerations. Statistical guidelines establish 0.5 as the critical minimum value for factor loadings. The statistical results presented in Table 1 reveal that most of the indicators passed the factor loading values, except for ORC5 and GCA1, which were excluded because their values were lower than the threshold of 0.7. A detailed examination of Table 1 reveals that the factor loadings demonstrated strong item reliability, with values ranging from 0.670–0.933, notably exceeding the conventional threshold of 0.5. GAIC4 (0.670) and TEC4 (0.550), while lower than other indicators, still maintained acceptable levels for retention in the measurement model. Cronbach’s alpha and RhoA values for all constructs exceeded the required 0.7 threshold. Internal consistency reliability was particularly strong, with Cronbach’s alpha values ranging from 0.804–0.924, and RhoA values spanning from 0.814–0.945, substantially exceeding the recommended 0.7 threshold. The highest reliability was observed in ENC (α = 0.907, RhoA = 0.945), followed by GEP (α = 0.924, RhoA = 0.924). The CR measurements also surpassed the established 0.70 benchmark, confirming internal validity requirements with values ranging from 0.863–0.946, demonstrating excellent construct reliability. Notably, GEP showed the highest CR value (0.946), followed by ENC (0.934) and GCA (0.926), indicating particularly strong internal consistency in these constructs. Note: ENC = Environmental Context, ORC = Organisational Context, TEC = Technological Context, GAIC = Generative AI capabilities, GET = Green Exploitative Innovation, GEP = Green Exploratory innovation, OCRE = Organisational Creativity, GCA = Green Competitive Advantage. Average Variance Extracted (AVE) calculations were conducted to assess convergent validity for each construct, with values exceeding 0.5 indicating acceptable convergent validity. The observed AVE values ranged from 0.560–0.814, with GEP demonstrating the strongest convergent validity (0.814), followed by GCA (0.758) and ENC (0.779). Even the constructs with lower AVE values, such as TEC (0.570) and OCRE (0.560), still met the minimum threshold, establishing adequate convergent validity across all constructs. 4.2. Discriminant validity. Two distinct methods were used to examine discriminant validity: the traditional Fornell – Larcker criterion and the more precise Heterotrait – Monotrait (HTMT) ratio analysis. Within the Fornell – Larcker matrix, diagonal values show the square roots of AVE for each construct, while off-diagonal values represent inter-construct correlation coefficients, as indicated in Table 2. Verifying discriminant validity requires comparing these AVE square roots against inter-construct correlations. The analysis confirmed that the square root of each construct’s AVE exceeded its correlations with other constructs, thereby fulfilling the Fornell – Larcker standards. The analysis confirmed robust discriminant validity, with the square root of AVE for each construct (ranging from 0.748–0.902) substantially exceeding their respective inter-construct correlations. The ENC demonstrated strong discriminant validity with a square root of AVE of 0.883, while GEP showed the highest discriminant validity with 0.902. Even constructs with relatively lower square roots of AVE, such as OCRE (0.748) and TEC (0.755), still maintained sufficient distinction from other constructs, thereby comprehensively fulfilling the Fornell – Larcker criterion. Supplemental validation emerged through HTMT ratio calculations, which provide a more rigorous methodological scrutiny. The HTMT matrix in Table 3 showed that measurements consistently fell below the strict 0.85 criterion, with all paired constructs displaying ratios of less than 0.90. These two validation methods together establish clear construct separation and theoretical distinction. The HTMT analysis revealed several notable relationship patterns among the constructs. The strongest HTMT ratios were observed between TEC and GAIC (0.835), OCRE and ENC (0.869), and OCRE and GCA (0.805), indicating substantial but acceptable relationships. Of particular interest was the high ratio between GET and GEP (0.841), suggesting a strong theoretical connection between exploitative and exploratory innovation dimensions while maintaining discriminant validity. The analysis also revealed notably low HTMT ratios between several construct pairs, such as ORC and OCRE (0.154), ORC and GEP (0.188), and ORC and GET (0.197), demonstrating clear construct distinction in these relationships. 4.3. Results of hypotheses. The theoretical model was analysed through sequential regression techniques implemented via SmartPLS software. The evaluation process centred on internal model calculations, with hypotheses tested by examining t-statistics and probability values. The complete statistical findings from this assessment are indicated in Table 4. According to the findings indicated in Table 4 and Figure 2. The TEC emerged as the strongest predictor of GAIC, demonstrating a significant positive relationship (β = 0.634, T value = 9.284), while the ORC showed a smaller but still significant effect (β = 0.182, T value = 2.368). Notably, the ENC failed to demonstrate a significant impact on GAIC (β = 0.016, T value = 0.298), suggesting that external environmental factors may play a less crucial role than initially theorised. Furthermore, GAIC was found to have a significant impact on OCRE (β = 0.223, T-value = 3.482) and GIA (β = 0.396, T-value = 6.190). The results also highlighted a particularly strong influence of OCRE on GCA (β = 0.646, T-value = 16.665), representing the strongest relationship in the model. The impact of GIA on GCA, while significant, showed a more moderate effect (β = 0.173, T-value = 3.083). These findings collectively support six out of seven hypothesised relationships, with all supported paths demonstrating strong statistical reliability in the observed relationships. Table 5 indicates the results for total effects. According to the findings in Table 5, TEC (β = 0.136, T-value = 3.697) and ORC (β = 0.038, T-value = 2.143) had a significant impact on GCA. On the other hand, ENC (β = 0.04, T-value = 0.272) did not have a statistically significant impact on GCA. Note: ENC = Environmental Context, ORC = Organisational Context, TEC = Technological Context, GAIC = Generative AI capabilities, GET = Green Exploitative Innovation, GEP = Green Exploratory innovation, OCRE = Organisational Creativity, GCA = Green Competitive Advantage Note: ENC = Environmental Context, ORC = Organisational Context, TEC = Technological Context, GAIC = Generative AI capabilities, GET = Green Exploitative Innovation, GEP = Green Exploratory innovation, OCRE = Organisational Creativity, GCA = Green Competitive Advantage. Table 5 indicates the results for total effects. The analysis of total effects revealed interesting patterns in the relationships between contextual factors and GCA. The TEC emerged as the strongest overall predictor of GCA, demonstrating a significant positive total effect (β = 0.136, T-value = 3.697). The ORC showed a smaller but still statistically significant total effect on GCA (β = 0.038, T-value = 2.143). On the other hand, the environmental context (ENC) did not have a statistically significant impact on GCA (β = 0.004, T-value = 0.272, p = 0.786). The findings suggest that organisations should prioritise TEC and ORC factors when developing strategies for achieving GCA. 5. Implications. 5.1. Comparison of results. The present study results showed that TEC, OC, and ENC were significantly associated with GAIC. This result is similar to an earlier study conducted by wael AL-khatib (wael AL-khatib 2023). Drawing upon the TOE framework, their study utilised an extensive online survey approach within the Jordanian retail industry. Their research findings revealed that key TOE elements – specifically relative advantage, top management support, organisational readiness, and customer pressures – demonstrated significant positive relationships with GAIC adoption. Their empirical evidence yielded strategic insights for technology solution providers, particularly regarding the development of structured implementation methodologies to enhance the effectiveness of technology deployment and promote the dissemination of innovation across organisations. Their research contributed practical frameworks for successful technology integration and innovation adoption within organisational settings. Furthermore, the current study revealed a significant association between GAIC and OCRE. This result can be compared with an earlier study by Mikalef and Gupta (2021). Based on the RBV, their research examined and categorised the technological resources that form organisational GAIC competencies. The study evaluated organisational technological capabilities by developing and validating measurement scales while exploring theoretical and empirical connections between these Note: ENC = Environmental Context, ORC = Organisational Context, TEC = Technological Context, GAIC = Generative AI capabilities, GIA = Green Innovation Ambidexterity, OCRE = Organisational Creativity, GCA = Green Competitive Advantage competencies, OCRE, and performance indicators. Their results substantiated a positive association between GAIC deployment and heightened OCRE, highlighting the transformational role of technological capabilities in enhancing creative organisational outcomes. Their methodological precision in scale development and theoretical foundation advanced the understanding of how technological competencies shape creative organisational processes. In addition, the current study found GAIC to be significantly associated with GIA. This result is compared to an earlier study by Wang and Zhang (2025). Their research explored how GAIC affects GIA among small and medium-sized enterprises. Their results revealed a significant positive correlation between GAIC and GIA, highlighting the role of technology in advancing sustainable innovation. The study underscored the importance of responsible technological governance structures in fostering sustainable organisational practices. Their work contributed to existing literature by identifying key mechanisms through which digital technologies promote sustainable business approaches in international commerce, specifically in cross-border digital trade. Additionally, the present study found that OCRE significantly impacts GCA. This result is somewhat similar to an earlier study by Musa and Enggarsyah. Their study examined the relationship between OCRE and GCA during periods of environmental disturbance. Cross-industry data analysis revealed that OCRE served primarily as a key driver in building organisational resilience amid disruptive conditions, whereas organisational agility showed stronger links to GCA. Their results highlighted the importance of developing a learning-focused organisational culture in strengthening OCRE and sustaining GCA. Lastly, the current study indicated a significant association between GIA and GCA. The result is similar to an earlier study by Mehmood et al.. Their study addressed a theoretical void by examining how GIA mediates the relationship between Big Data Note: ENC = Environmental Context, ORC = Organisational Context, TEC = Technological Context, GCA = Green Competitive Advantage Analytics (BDA) and GCA. Their analysis of manufacturing sector data revealed that BDA had a substantial impact on GIA, which subsequently served as a significant mediator in the relationship between BDA and GCA. Hence, a significant association between GIA and GCA was indicated. 5.2. Theoretical implications of research. This research makes several substantive theoretical contributions to the extant literature. Primarily, it extends the TOE framework by incorporating GAIC within the technological context, thereby enhancing scholarly understanding of contemporary technological influences on organisational outcomes. While extant literature has primarily concentrated on traditional technological innovations, this research presents a novel examination of GAIC through the TOE perspective, providing refined theoretical insights into organisational technology assimilation and utilisation processes. Furthermore, this investigation contributes to the OCRE literature by establishing theoretical linkages between GAIC and OCRE. Through a systematic examination of the mechanisms that facilitate enhanced creative processes, this research extends beyond conventional human-centred creativity paradigms to encompass human-AI collaborative creativity. This theoretical progression addresses contemporary scholarly discourse regarding technological enhancement of creative processes in organisational settings, contributing to the emerging theoretical understanding of hybrid intelligence. Additionally, this research advances ambidexterity theory through empirical validation of GIA constructs. By examining organisational pursuit of concurrent GET and GEP through GAIC, this study extends organisational ambidexterity theory into environmental sustainability contexts. This theoretical synthesis provides scholarly frameworks for understanding organisational management of competing demands in pursuit of environmental innovation. Moreover, this investigation advances RBV by examining the transformation of GAIC into GCA. The findings demonstrate how interactions between technological capabilities and organisational processes generate distinctive, inimitable resources that enhance long-term GCA. This theoretical insight expands current understanding of digital resource contributions to organisational environmental performance. Lastly, this research contributes to the innovation literature by proposing and validating an integrated theoretical framework that connects GAIC, OCRE, and GIA. Through empirical validation of these relationships, this study establishes theoretical foundations for understanding how organisations can leverage advanced technologies to achieve economic and environmental objectives. This integrated theoretical perspective advances a comprehensive understanding of sustainable innovation processes in the digital age. 5.3. Practical implications of research. This research presents several significant implications for managers, practitioners, and policymakers leveraging GAIC for sustainable innovation outcomes. For organisational managers, the findings highlight the strategic importance of developing comprehensive GAIC implementation frameworks that extend beyond basic adoption processes. Managers should prioritise establishing organisational environments that facilitate the integration of GAIC with existing creative processes. Specifically, the results indicate that organisations should invest concurrently in technical infrastructure and human capital development, as the synergy between these elements substantially enhances creative output and innovation performance. For practitioners in manufacturing industries, this study offers pragmatic guidelines for balancing GET and GEP. The findings demonstrate that successful GAIC implementation necessitates systematic attention to short-term efficiency gains and long-term innovative capabilities. Practitioners should establish precise metrics to evaluate the impact of GAIC on both incremental improvements and radical innovations in their sustainability initiatives. Furthermore, this research suggests that organisations should develop specialised training programmes to enhance employees’ ability to collaborate effectively with GAIC systems, particularly emphasising areas where human creativity and GAIC can create complementary advantages. From a policy perspective, the research findings present substantial implications for governmental bodies and regulatory authorities. The research suggests that policymakers should develop frameworks that promote the adoption of GAIC while ensuring its responsible and sustainable implementation. This encompasses creating incentive structures that reward organisations for utilising GAIC to advance green innovation initiatives, as well as establishing guidelines for the ethical deployment of AI in sustainability contexts. Additionally, policymakers should consider developing standards for measuring and reporting the environmental impact of GAIC-driven innovations, enabling organisations to benchmark their performance and identify opportunities for improvement. For industry leaders and executives, this research highlights the importance of fostering a culture that values both technological innovation and environmental responsibility. The findings suggest that organisations should establish cross-functional teams dedicated to identifying and implementing GAIC-driven sustainability initiatives, while developing mechanisms for sharing knowledge and disseminating best practices across organisational units. Moreover, executives should prioritise investments in GAIC, specifically targeting environmental challenges, as the results demonstrate that these investments yield both operational and strategic benefits. 5.4. Research limitations and future research directions. Despite its contributions, this study has several limitations that provide opportunities for future research. First, the cross-sectional nature of our data collection limits the ability to capture the dynamic evolution of GAIC implementation and its long-term effects on organisational creativity and green innovation outcomes. Future longitudinal studies could examine how organisations’ GAIC develop over time and investigate the temporal aspects of sustainable innovation development. Such research could provide valuable insights into the causal relationships between the adoption of GAIC and organisational performance across different stages of implementation. Second, while the sample focused on manufacturing industries in Taiwan, the generalisability of these findings to other contexts and geographical regions may be limited. Cultural factors and institutional environments specific to Taiwan might influence the relationships. Future research could adopt a cross-cultural perspective by replicating this study across different countries and industries, particularly in developed economies where GAIC adoption patterns and environmental regulations may differ. Additionally, comparative studies between emerging and developed economies could reveal important contextual factors affecting GAIC implementation and its outcomes. Third, this research primarily examined organisational-level outcomes, potentially overlooking important individual and team-level dynamics in the adoption and usage of GAIC. Future studies could adopt a multilevel approach to investigate how individual attitudes, team dynamics, and organisational factors collectively influence the success of GAIC implementation. Particularly valuable would be research examining how different leadership styles and team compositions influence the effectiveness of GAIC in driving sustainable innovation. 6. Conclusion. This research advances both theoretical understanding and practical applications by examining how GAIC is implemented in manufacturing organisations through the TOE framework. This research analysis reveals that while TEC and ORC dimensions significantly shape GAIC implementation, ENC factors demonstrate no substantial influence. The findings establish GAIC’s positive influence on OCRE and GIA, which subsequently strengthens GCA. This study expands the scope of the TOE framework by incorporating GAIC, thereby enriching our understanding of human-AI collaborative creativity. It extends organisational ambidexterity theory into the domain of environmental sustainability. For manufacturing practitioners, the findings of this study provide strategic guidance for implementing GAIC, emphasising the critical balance between developing technological infrastructure and organisational preparedness. Nevertheless, this study’s contextual boundaries, particularly its focus on Taiwanese manufacturing firms, present certain limitations regarding generalisability. Future scholarly endeavours could address these constraints through cross-cultural comparative analyses, longitudinal investigations of the effects of GAIC implementation, and an examination of potential moderating variables. Disclosure statement No potential conflict of interest was reported by the author(s). Data availability statement The data presented in this study are available on request from the corresponding author.