Mediators and Moderators in AI Capability Models

Selecting appropriate mediating and moderating variables in AI capability research is one of the most important decisions scholars face when designing structural equation models (SEM). Empirical evidence consistently shows that artificial intelligence capability rarely exerts a simple, unconditional direct effect on firm performance. Instead, AI capability generates competitive advantage indirectly by activating mediating organizational capabilities—such as agility, resilience, absorptive capacity, and innovation ambidexterity—while its effectiveness is conditioned by internal cultural factors and external environmental turbulence. This article synthesizes the mediating and moderating variables tested across peer-reviewed studies to help researchers construct rigorous SEM frameworks.

Mediators and Moderators in AI Capability Models

Part 1: Why Direct-Effects Models Are Insufficient in AI Capability Research

In early information systems studies, researchers frequently tested direct linear paths from technology investments to financial returns, yielding mixed or non-significant findings. Grounded in Dynamic Capabilities Theory (Teece, 2007) and the Resource-Based View (Barney, 1991; Mikalef & Gupta, 2021), contemporary scholarship recognizes that firm-level AI capability must be embedded into dynamic organizational routines to create value. In their systematic review of 35 empirical studies, Nguyen, T. P., Tran, D. K., & Nguyen, T. H. H. (2026) document that 28 of the 35 studies (80%) incorporate explicit mediating variables and 24 of the 35 studies (68.6%) incorporate moderating boundary conditions (Table 5 of the review), confirming that mediated and moderated-mediation architectures represent the empirical norm in this field.

Part 2: Four Families of Mediating Variables in AI Capability Models (28 Studies)

When specifying mediating and moderating variables in AI capability research, scholars can select from four empirically validated families of mediators depending on their dependent variable:

1. Agility and Strategic Flexibility Mediators

Organizational agility, decision-making agility, adaptive capability, and strategic flexibility represent the most widely used mediating mechanisms linking AI capability to firm performance (Lee, Lee, & Kim, 2026; Lu, Fan, Troise, & Bresciani, 2026; T. L. Le, 2026; Zhong, Ady, & Indrawati, 2026). By accelerating real-time market sensing and reducing decision latency, AI capability enhances a firm’s ability to pivot operations and seize market opportunities rapidly.

2. Supply Chain and Organizational Resilience Mediators

In operations, logistics, and crisis management studies, scholars model supply chain resilience, employee resilience, financial resilience, and strategic resilience as key mediators (Ratanacharooenchai & Jantapoon, 2026; Javed, Švecová, Danko, & Tučková, 2026; Al Halbusi, Popa, Soto-Acosta, & Abdelallag, 2025; Wang & Zhang, 2026). These constructs explain how AI-driven visibility and predictive risk analytics enable organizations to absorb external shocks and sustain performance.

3. Innovation Ambidexterity and Business Model Transformation Mediators

When the target outcome is digital transformation, servitization, or corporate sustainability, researchers frequently specify innovation ambidexterity (exploratory and exploitative innovation), green ambidexterity, open innovation, or business model transformation as mediators (Ahmad, 2025; Al-Khatib, Al-Fadhel, & Khattab, 2026; Duong, 2026; A. D. Le, Vu, Phan, & Dao, 2026).

4. Knowledge Integration and Resource Orchestration Mediators

Drawing on the Knowledge-Based View, several studies model green knowledge integration, absorptive capacity, data-driven culture, and resource orchestration as mediating conduits that translate raw algorithmic outputs into organizational learning and sustainable performance (Fosso Wamba, Queiroz, Pappas, & Sullivan, 2024; Chen, Li, Tang, Li, & Guo, 2026).

Mediators and Moderators in AI Capability Models

Part 3: Internal and External Moderating Variables (24 Studies)

To explain when AI capability exerts a stronger or weaker impact on mediators and outcomes, empirical researchers integrate two categories of moderating variables:

  • External Environmental and Institutional Moderators: Environmental dynamism / turbulence (Dubey et al., 2020; Li, Xu, Zhang, Liu, & Li, 2025; Lee et al., 2026), technological turbulence (A. D. Le et al., 2026), government support and regulatory intervention, and geopolitical shocks (AlQershi, Ramayah, Alzuraiki, & Emam, 2026). Under high turbulence, the information-processing and sensing advantages of AI capability become substantially more valuable.
  • Internal Organizational and Socio-Technical Moderators: Absorptive capacity (Abou-Foul, Ruiz-Alba, & López-Tenorio, 2023), green knowledge management (Kumar, Kumar, Chaudhuri, Chatterjee, & Vrontis, 2025), cybersecurity awareness (Chowdhury, Ren, & Dey, 2026), corporate foresight (Camargo, Morales, Nascimento, & Tortorella, 2025), and leadership and policy environment (Zhong et al., 2026).
Variable CategorySpecific Constructs Tested in the CorpusTheoretical JustificationRepresentative Empirical Studies (Nguyen et al., 2026) 
Mediators: Agility & AdaptationOrganizational agility, decision-making agility, strategic flexibility, adaptive capabilityDynamic Capabilities Theory (speed of sensing and seizing)Lee et al. (2026); Lu et al. (2026); T. L. Le (2026); Zhong et al. (2026)
Mediators: ResilienceSupply chain resilience, employee resilience, financial and strategic resilienceDynamic Capabilities Theory & RBV (shock absorption and recovery)Ratanacharooenchai & Jantapoon (2026); Javed et al. (2026); Al Halbusi et al. (2025)
Mediators: Innovation & KnowledgeInnovation ambidexterity, green ambidexterity, business model transformation, data-driven cultureAmbidexterity Theory & Knowledge-Based View (KBV)Ahmad (2025); Al-Khatib et al. (2026); Duong (2026); Fosso Wamba et al. (2024)
Moderators: External ContingenciesEnvironmental dynamism, technological turbulence, regulatory pressure, geopolitical shocksContingency Theory & Institutional TheoryDubey et al. (2020); Li et al. (2025); A. D. Le et al. (2026); AlQershi et al. (2026)
Moderators: Internal CapabilitiesAbsorptive capacity, green knowledge management, cybersecurity awareness, corporate foresightKBV, Resource Orchestration, & Complementarity TheoryAbou-Foul et al. (2023); Kumar et al. (2025); Chowdhury et al. (2026); Camargo et al. (2025)

Part 4: Specifying Moderated Mediation Models in PLS-SEM and CB-SEM

Across the 35 reviewed studies, researchers predominantly estimate these relationships using Partial Least Squares SEM (PLS-SEM in 25 studies), Covariance-Based SEM (CB-SEM in 9 studies), or PROCESS Macro (5 studies), increasingly supplementing symmetric regression paths with Fuzzy-set Qualitative Comparative Analysis (fsQCA in 4 studies) to test how multiple moderators and mediators combine configurationally. Importantly, 13 studies in the corpus also test AI capability itself as a mediator, moderator, or endogenous construct driven by upstream antecedents such as green strategic intent or intellectual capital (Zahoor, Usman, Khalil, Alwehabie, & Yasin, 2025; Lu et al., 2026). Researchers designing new structural models can review the complete coding of mediating and moderating mechanisms in AI capability and dynamic capabilities research in Nguyen, T. P., Tran, D. K., & Nguyen, T. H. H. (2026), AI capability in dynamic capabilities: A systematic literature review using the TCCM framework, in Proceedings of the International Conference on Business Sustainability (ICBS 2026) (pp. 357–370), Information and Communications Publishing House.

Mediators and Moderators in AI Capability Models

References (APA 7th Edition)

Abou-Foul, M., Ruiz-Alba, J. L., & López-Tenorio, P. J. (2023). The impact of artificial intelligence capabilities on servitization: The moderating role of absorptive capacity—A dynamic capabilities perspective. Journal of Business Research, 157, 113609. https://doi.org/10.1016/j.jbusres.2022.113609

Ahmad, Z. (2025). Unlocking AI capabilities: Exploring strategic fit, innovation ambidexterity, and digital entrepreneurial intent in driving digital transformation. Journal of Management Development, 44(2), 194–218. https://doi.org/10.1108/JMD-06-2024-0197

Al Halbusi, H., Popa, S., Soto-Acosta, P., & Abdelallag, M. (2025). The nexus of managerial and technical AI knowledge, disruptive innovation and the circular economy: The role of organizational change capability and financial resilience. Technology in Society, 82, 102937. https://doi.org/10.1016/j.techsoc.2025.102937

Al-Khatib, A. W., Al-Fadhel, H. A., & Khattab, M. (2026). How can generative artificial intelligence improve digital supply chain performance in manufacturing firms? Analyzing the mediating role of innovation ambidexterity using hybrid analysis through CB-SEM and PLS-SEM. Knowledge Management Research & Practice, 24(1), 1–16.

AlQershi, N., Ramayah, T., Alzuraiki, M., & Emam, A. S. M. (2026). Artificial intelligence and strategic resilience in international business: A new theory of multinational adaptation to geopolitical shocks. Review of International Business and Strategy, 36(1), 1–27. https://doi.org/10.1108/RIBS-03-2025-0061

Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120.

Bibi, M., Tan, T. G., & Yau, H. (2025). Exploring the impact of AI capabilities on employee well-being: A mediated moderation analysis. SAGE Open, 15(3), 21582440251365692. https://doi.org/10.1177/21582440251365692

Camargo, G. F., Morales, V. M., Nascimento, D. L., & Tortorella, G. L. (2025). Exploring the role of corporate foresight in the coordination between dynamic capabilities and artificial intelligence in green innovation: A multi-continent comparison. IEEE Transactions on Engineering Management, 72, 1–14. https://doi.org/10.1109/TEM.2025.3533456

Chen, R., & Li, Z. (2026). Empirical analysis of the roles of dynamic sustainable capabilities and artificial intelligence in accelerating circular economy-driven innovation: Insights from Chinese manufacturing firms. Technological Forecasting and Social Change, 224, 124489. https://doi.org/10.1016/j.techfore.2025.124489

Chen, X., Li, Y., Tang, G., Li, W., & Guo, Z. (2026). Exploring how artificial intelligence capabilities impact corporate sustainability performance: Insights from Chinese manufacturing firms. Technovation, 151, 103428. https://doi.org/10.1016/j.technovation.2025.103428

Chowdhury, S., Ren, S., & Dey, P. K. (2026). AI assimilation for sustainable and resilient supply chains: Role of green ambidexterity and cybersecurity awareness. Technovation, 153, 103562. https://doi.org/10.1016/j.technovation.2026.103562

Dubey, R., Gunasekaran, A., Childe, S. J., Bryde, D. J., Giannakis, M., Foropon, C., & Hazen, B. T. (2020). Big data analytics and artificial intelligence pathway to operational performance under the effects of entrepreneurial orientation and environmental dynamism: A study of manufacturing organisations. International Journal of Production Economics, 226, 107599. https://doi.org/10.1016/j.ijpe.2019.107599

Duong, C. D. (2026). The green transition journey: How digital platforms and AI capabilities drive green ambidexterity innovation, circular economy, and sustainable performance. World Development Sustainability, 8, 100268. https://doi.org/10.1016/j.wds.2025.100268

Fosso Wamba, S., Queiroz, M. M., Pappas, I. O., & Sullivan, Y. (2024). Artificial intelligence capability and firm performance: A sustainable development perspective by the mediating role of data-driven culture. Information Systems Frontiers, 26(6), 2315–2330. https://doi.org/10.1007/s10796-023-10456-3

Hart, S. L. (1995). A natural-resource-based view of the firm. Academy of Management Review, 20(4), 986–1014.

Javed, M., Švecová, L., Danko, L., & Tučková, Z. (2026). Harnessing artificial intelligence and employee resilience for enhanced business performance: A resource-based and dynamic capabilities perspective. Human Behavior and Emerging Technologies, 2026, 1701205. https://doi.org/10.1155/hbet/1701205

Kugeza, G. (2026). AI-augmented decision-making and sustainability performance in SMEs: A hybrid PLS-SEM, ANN, and multi-group analysis. Business Strategy and the Environment, 35(2), 812–829.

Kumar, S., Kumar, V., Chaudhuri, R., Chatterjee, S., & Vrontis, D. (2025). AI capability and environmental sustainability performance: Moderating role of green knowledge management. Technology in Society, 81, 102878. https://doi.org/10.1016/j.techsoc.2025.102878

Le, A. D., Vu, N. T., Phan, T. T. H., & Dao, V. T. (2026). AI-enabled business model transformation and entrepreneurial quality: The moderating effect of technological turbulence. Technology in Society, 85, 103062. https://doi.org/10.1016/j.techsoc.2025.103062

Le, T. L. (2026). Green AI, strategic flexibility, and open innovation: Advancing circular economy and sustainable performance in SMEs. Journal of Open Innovation: Technology, Market, and Complexity, 12(1), 100452.

Lee, K., Lee, S., & Kim, S. (2026). AI for decision-making: Exploring the linkage from AI capability to firm performance. Aslib Journal of Information Management, 78(3), 512–534. https://doi.org/10.1108/AJIM-04-2025-0241

Li, L., Xu, C., Zhang, Q., Liu, Y., & Li, Q. (2025). Leveraging generative AI capabilities for competitive advantage in B2B service firms: A moderated mediation analysis of environmental dynamism and digital service innovation. Industrial Marketing Management, 124, 112–126.

Lu, J., Fan, L., Troise, C., & Bresciani, S. (2026). Exploring the nexus between artificial intelligence capability, multidimensional intellectual capital, and organizational agility of SMEs. Journal of Innovation & Knowledge, 11(2), 100697. https://doi.org/10.1016/j.jik.2026.100697

Mikalef, P., & Gupta, M. (2021). Artificial intelligence capability: Conceptualization, measurement calibration, and empirical study on its impact on organizational creativity and firm performance. Information & Management, 58(3), 103434. https://doi.org/10.1016/j.im.2021.103434

Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. BMJ, 339, b2535. https://doi.org/10.1136/bmj.b2535

Nguyen, T. P., Tran, D. K., & Nguyen, T. H. H. (2026). AI capability in dynamic capabilities: A systematic literature review using the TCCM framework. In Proceedings of the International Conference on Business Sustainability (ICBS 2026) (pp. 357–370). Information and Communications Publishing House. https://www.researchgate.net/publication/416933536_AI_CAPABILITY_IN_DYNAMIC_CAPABILITIES_A_SYSTEMATIC_LITERATURE_REVIEW_USING_THE_TCCM_FRAMEWORK

Paul, J., & Rosado-Serrano, A. (2019). Gradual internationalization vs born-global/international new venture models: A review and research agenda. International Marketing Review, 36(6), 830–858. https://doi.org/10.1108/IMR-10-2018-0280

Ratanacharooenchai, C., & Jantapoon, K. (2026). The mediating role of supply chain resilience in the relationship between AI capabilities and sustainability performance: Evidence from manufacturing SMEs. Sustainable Futures, 11, 101789. https://doi.org/10.1016/j.sftr.2026.101789

Teece, D. J. (2007). Explicating dynamic capabilities: The nature and microfoundations of (sustainable) enterprise performance. Strategic Management Journal, 28(13), 1319–1350. https://doi.org/10.1002/smj.640

Zahoor, N., Usman, M., Khalil, S., Alwehabie, A., & Yasin, N. (2025). Green strategic intent, artificial intelligence capability, and behavioral dynamics of achieving circular economy goals in the public sector. Technological Forecasting and Social Change, 215, 124062. https://doi.org/10.1016/j.techfore.2025.124062

Lên đầu trang