Artificial intelligence (AI) is increasingly adopted by small and micro-enterprises to enhance international competitiveness. However, limited research examines how internal AI capability and external AI tool utilization jointly shape export performance. Drawing on the resource-based view and digital resource configuration perspective, this study conceptualizes internal AI capability and external AI tool utilization as distinct but potentially overlapping AI-related resources. Using survey data from 475 exporting small and micro-enterprises in Yiwu International Trade City, we conduct regression analyses to investigate the individual and interactive effects of these two AI-related resources on export performance. The results indicate that both internal AI capability and external AI tool utilization positively affect export performance. Importantly, their interaction is negative and significant, suggesting diminishing marginal returns when both resources are highly developed. This finding indicates that overlapping AI-related investments may reduce each resource’s incremental contribution under resource-constrained conditions. By clarifying how internally developed AI capability and externally accessed AI tools interact in export settings, this study advances understanding of digital resource configuration and provides practical guidance for AI-related investment decisions in small firms. 1. Introduction Small and micro-enterprises provide an important context for examining the relationship between AI utilization and export performance for two main reasons. First, small and micro-enterprises have become increasingly important participants in international trade, particularly in export-oriented economies such as China. For these firms, export performance is not only a financial outcome but also a critical determinant of firm survival, competitiveness, and growth in international markets [ 8, 9]. Despite their relatively limited organizational scale, export-oriented small and micro-enterprises play an important role in regional economic activity and cross-border trade networks. Second, the growing diffusion of AI technologies has increasingly enabled small firms to adopt digital tools to support export-related activities [ 10, 11]. AI-enabled applications such as multilingual translation, customer communication, content generation, and overseas market information retrieval can help small firms reduce communication barriers and improve operational efficiency in export markets [ 12]. Recent research further suggests that digital technologies and digital platforms increasingly shape how small firms participate in international markets and improve export-related competitiveness [ 13]. However, as the availability of AI-related technologies continues to expand, small and micro-enterprises also face increasingly important strategic decisions regarding how to allocate limited financial and managerial resources across different forms of AI-related investments [ 10, 14]. Because these firms are typically unable to invest heavily in every available AI-related technology or capability, questions regarding the most effective configuration of AI-related resources become increasingly important. Addressing this issue is particularly important because small and micro-enterprises must strategically prioritize AI-related investments under conditions of limited organizational resources. Within this stream of research, two related but conceptually distinct AI-related assets are particularly relevant at the firm level. Internal AI capability captures the firm’s organizational capacity to select, integrate, govern, and routinize AI-related solutions through data readiness, employee skills, process redesign, and managerial practices, so that AI use becomes embedded in export-related workflows. In contrast, external AI tool utilization refers to the intensity with which a firm relies on platform-provided AI applications that are accessed as standardized services (e.g., AI translation, automated customer communication, content generation, and AI-assisted market information search) to support export activities. These tools are typically ready to use and do not require firms to develop algorithms or infrastructure in-house [ 15]. While learning by using external tools can contribute to capability accumulation over time, the two constructs remain analytically distinct. One reflects reliance on externally sourced AI services, whereas the other reflects an internally embedded ability to deploy AI effectively within organizational routines. Although prior research suggests the possibility of both complementarity and overlap between externally accessed digital solutions and internally developed capabilities, we argue that partial substitution is more likely to emerge in export environments characterized by resource constraints. In such contexts, both internally developed AI capability and externally provided AI tools increasingly support similar export activities, and their functions in export operations may increasingly overlap. Under these conditions, the marginal contribution of one capability is likely to decline as the other becomes more developed. Therefore, rather than generating purely additive performance gains, the interaction between the two may exhibit partial substitution and diminishing marginal returns. To examine this argument, we focus on exporting small and micro-enterprises in Yiwu International Trade City, a major export cluster where platform-based trade and externally provided AI services are widely accessible. Using survey data from 475 exporting firms, we estimate multiple regression models to assess the direct and interaction effects of external AI tool utilization and internal AI capability on export performance. The results indicate that both external AI tool utilization and internal AI capability are positively associated with export performance. At the same time, their interaction effect is negative, suggesting that when one resource is already highly developed, the incremental contribution of the other becomes smaller. These findings provide evidence of partial substitution between overlapping digital resources in resource-constrained export contexts. Understanding this configuration logic is also relevant to sustainability because inefficient or overlapping digital investments may place additional financial and managerial burdens on small firms, whereas better-aligned AI adoption can support more sustainable export growth. This study contributes to the literature in three ways. First, this study advances the literature on digital resource configuration by identifying conditions under which internal AI capability and external AI tool utilization exhibit partial substitution rather than complementary effects. Second, by focusing on exporting small and micro-enterprises operating in a highly platformized trade environment, the study extends digital resource configuration research to contexts characterized by limited capital and managerial attention. Third, the findings provide practical implications for sustainable digital investment strategies, suggesting that small and micro-enterprises should carefully balance internal capability development and external tool adoption to avoid inefficient resource duplication and support sustainable export performance improvement. 2. From Traditional Trade to Platform-Based AI Adoption: The Yiwu Context More importantly, Yiwu has moved beyond traditional offline trade and basic digitalization toward platform-based AI adoption. Digital trade platforms such as Chinagoods have introduced AI-enabled services to support small-commodity export activities, including multilingual translation, product-content generation, visual content creation, digital marketing, customer communication, and trade matching. Unlike general-purpose AI tools, these platform-based AI applications are embedded in Yiwu’s small-commodity export context and are directly connected to routine export tasks such as product display, cross-language communication, customer inquiry handling, online promotion, and marketing material preparation. Consequently, AI technologies in Yiwu are not merely experimental digital tools but increasingly constitute operational resources that firms use in their day-to-day export activities. This context provides a suitable empirical setting for the present study. Fieldwork conducted in July 2024 indicated that small and micro exporting firms in Yiwu varied considerably in their use of platform-provided AI tools. Some firms actively used AI tools for translation, content generation, short-video production, customer communication, and digital marketing, whereas others used them only occasionally or in a limited manner. At the same time, firms also differed in their internal AI capability, including digital infrastructure, employee AI-related skills, organizational learning capacity, and managerial support for integrating AI into export-related workflows [ 17, 22]. Such variation in both external AI tool utilization and internal AI capability makes Yiwu an appropriate context for examining how externally provided AI tools and firms’ internal AI capability jointly shape export performance. 3. Theory and Hypotheses 3.1. External AI Tool Utilization and Export Performance H1:External AI tool utilization positively affects firm export performance. 3.2. Internal AI Capability and Export Performance H2:Internal AI capability positively affects firm export performance. 3.3. External–Internal Resource Interaction and Export Performance Simultaneous investments in both resources can also generate duplicated task coverage, additional coordination, monitoring, and learning costs, as well as heavier demands on limited managerial attention [ 45]. This tendency may be especially salient in small exporting firms with constrained financial and organizational resources, since committing scarce resources to both internal AI capability building and extensive external AI tool utilization can crowd out alternative export enhancing investments, such as customer acquisition, product adaptation, logistics improvement, and relationship building in foreign markets [ 46]. While complementarity may still arise in more customized or nonroutine export activities, we argue that partial substitution is more likely in this context, and thus the interaction effect between external AI tool utilization and internal AI capability on export performance is expected to be negative. Accordingly, we hypothesize: H3:The interaction between external AI tool utilization and internal AI capability negatively affects firm export performance. Figure 1 presents the research model of the study and summarizes the hypothesized relationships among the key variables. Specifically, the model illustrates the positive effects of external AI tool utilization and internal AI capability on export performance, as well as the negative interaction effect between the two AI-related resources. 4. Data and Method 4.1. Sample and Research Context This study examines exporting small and micro-enterprises operating in Yiwu International Trade City, one of China’s largest clusters of export-oriented small-commodity businesses [ 19, 20]. In recent years, AI-based applications tailored to small merchants, such as AI translation tools, digital trade assistants, and automated content-generation services, have been increasingly promoted in Yiwu. These externally provided tools are designed to facilitate cross-border trade by supporting routine export-related tasks such as translation, product-content generation, customer communication, and online promotion. At the same time, firms differ substantially in their ability to use and integrate these tools into daily export activities. The coexistence of widespread external tool availability and heterogeneous internal capability conditions makes Yiwu an appropriate setting for examining how external AI tool utilization and internal AI capability jointly influence export performance. To support questionnaire development and better understand the research context, we conducted preliminary field visits and interviews with exporting merchants. The interviews suggested that merchants frequently relied on platform-provided external AI tools for routine export-related tasks, such as translation, product-description generation, and customer inquiry handling, while firms differed in their ability to integrate these tools into daily operations. Interviewees also expressed differing views on whether these external AI tools substituted for or strengthened internal AI capability development. These insights helped refine the questionnaire items and informed the distinction between external AI tool utilization and internal AI capability. Prior to the survey, all respondents were fully informed of the academic purpose, voluntary participation, and confidentiality of the study both orally and in writing. This study was conducted in strict compliance with academic ethical norms and the principles of informed consent. Respondents were assured that their answers would be used solely for academic research and reported only in aggregate form, and no personally identifiable information was disclosed throughout the research process to protect participants’ legitimate rights and privacy. A total of 521 questionnaires were distributed. After excluding responses from firms primarily focused on domestic sales and questionnaires with substantial missing values on key variables, 475 valid observations were retained for analysis. 4.3. Common Method Bias Before conducting hypothesis testing, we assessed potential common method bias using both Harman’s single-factor test and a common latent factor approach. The Harman single-factor test showed that the first unrotated factor explained 43.11% of the total variance, below the commonly used threshold of 50%, suggesting that common method bias is unlikely to be severe. In addition, a common latent factor was introduced into the CFA model to further assess the potential influence of common method variance. The inclusion of the common latent factor did not significantly improve model fit (Δχ 2 = 0.05, p = 0.822), and the common-method loadings were not statistically significant. These results suggest that common method bias is unlikely to be a serious concern in this study. 4.4. Analytical Strategy This study employs ordinary least squares (OLS) regression to test the proposed hypotheses. The dependent variable and the two focal explanatory variables were operationalized as composite measures based on multiple five-point Likert scale items. OLS regression was adopted because the study aims to examine both the direct effects of external AI tool utilization and internal AI capability on export performance, as well as the interaction effect between these two variables. Since the analysis focuses on testing direct and interactive relationships among composite constructs rather than estimating a full latent variable structural model, OLS provides a parsimonious and interpretable analytical approach. To construct the interaction term, external AI tool utilization and internal AI capability were mean-centered prior to multiplication. Mean-centering also facilitates coefficient interpretation and helps reduce nonessential multicollinearity in interaction models. In OLS regression, the normality assumption primarily concerns the residuals rather than the raw survey variables themselves. To assess this assumption, we examined the standardized residuals from the full regression model. The residual distribution was centered close to zero and approximated normality without evidence of serious distortion. Visual inspection of the histogram and normal Q-Q plot further indicated only minor deviations from normality. Given the composite nature of the scale measures, the sample size of 475 firms, and the study’s focus on direct and interaction effects, OLS regression provides an appropriate and interpretable method for analyzing the survey data. 5. Results 5.1. Descriptive Statistics and Correlations Table 2 reports the descriptive statistics and correlation matrix for all variables. External AI tool utilization is positively correlated with export performance (r = 0.388, p < 0.05), and internal AI capability is also positively correlated with export performance (r = 0.439, p < 0.05). In addition, external AI tool utilization and internal AI capability are positively correlated with each other (r = 0.431, p < 0.05). To assess potential multicollinearity, variance inflation factor (VIF) tests were conducted. The VIF values range from 1.02 to 1.28, with a mean of 1.08, well below conventional thresholds, indicating that multicollinearity is not a concern. 5.2. Results of Hypothesis Testing Table 3 reports the results of the OLS regression analyses. Model (1) includes only the control variables. Models (2) and (3) separately introduce external AI tool utilization and internal AI capability, respectively, to examine their individual associations with export performance. In Model (2), external AI tool utilization shows a positive and significant effect on export performance (β = 0.364, p < 0.01). Similarly, Model (3) indicates that internal AI capability is positively and significantly associated with export performance (β = 0.447, p < 0.01). Model (4) includes both external AI tool utilization and internal AI capability simultaneously. Both variables remain positive and significant (β = 0.228, p < 0.01; β = 0.345, p < 0.01, respectively), providing support for H1 and H2. These results suggest that both external AI tool utilization and internal AI capability independently contribute to export performance. Model (5) introduces the interaction term between external AI tool utilization and internal AI capability to test H3. The interaction coefficient is negative and statistically significant (β = –0.107, p < 0.01), supporting H3. This finding suggests that the positive effect of one digital resource on export performance diminishes as the level of the other resource increases. Figure 2 further illustrates this interaction. When internal AI capability is low, external AI tool utilization exhibits a stronger positive association with export performance. In contrast, when internal AI capability is high, the positive effect of external AI tool utilization on export performance becomes weaker. This finding suggests that firms possessing stronger internal AI capability may rely less on external AI tools to achieve export performance gains. The interaction pattern therefore indicates a partial substitution relationship between the two digital resources rather than a purely complementary effect. 5.3. Robustness Checks First, given that a small number of firms in the sample reported more than 200 employees, the analyses were re-estimated after excluding these firms, and the results remained substantively unchanged. In addition, the models were re-run after excluding firms with more than 50 employees to focus more strictly on small and micro-enterprises. The main and interaction effects remained consistent in both direction and statistical significance. Second, several control variables were operationalized as dummy variables to preserve degrees of freedom. To ensure robustness, the regressions were re-estimated using their original categorical classifications, and the results remained substantively unchanged. 6. Discussion 6.1. Theoretical Implications Using survey data from 475 exporting small and micro-enterprises in Yiwu International Trade City, this study examined how external AI tool utilization and internal AI capability jointly shape export performance. The results show that both external AI tool utilization and internal AI capability are positively associated with export performance. This finding is broadly consistent with prior studies suggesting that digitalization and digital resources can improve firm performance and internationalization outcomes [ 6, 52]. Research on digital platforms and SME internationalization has similarly suggested that externally provided digital resources can help small firms overcome resource constraints and participate more effectively in international markets [ 8, 16, 53]. Recent studies have further argued that digital platforms can reshape organizational capabilities and facilitate capability reconfiguration among SMEs [ 54]. Consistent with this view, our findings suggest that both externally accessed AI tools and internally developed AI capability represent important AI-related resources through which small and micro-enterprises can enhance export performance. The findings also clarify the roles of external AI tools and internal AI capability in export settings. External AI tools provide firms with access to ready-made AI functionalities through platforms and service providers, whereas internal AI capability reflects a firm’s ability to integrate and utilize AI within its own operations and routines. Although these two resources differ in how AI-related support is accessed and deployed, both can be applied to similar export-related activities, such as translation, customer communication, content generation, and market information processing. 6.2. Practical Implications From a practical standpoint, the findings suggest that exporting small and micro-enterprises should approach digital investment decisions with greater strategic alignment. While both internal AI capability development and external AI tool utilization independently enhance export performance, simultaneously intensifying both may not always be efficient. Firms with limited resources should carefully evaluate how internally developed AI capability and externally provided AI tools can be balanced. While external AI tools may provide accessible and cost-efficient support for export activities, firms with stronger internal AI capability may derive less incremental benefit from extensive reliance on standardized external solutions. Careful calibration between internal capability accumulation and external tool adoption may therefore support more efficient and sustainable export performance improvement by reducing redundant digital investment and improving the long-term viability of AI-enabled exporting. 6.3. Limitations and Future Research This study has several limitations that offer directions for future research. First, the data are drawn from exporting small and micro-enterprises located in Yiwu, China. Although Yiwu represents a highly relevant platform-based export environment, the regional and industry concentration may limit the generalizability of the findings. In addition, this study adopts a context-specific operationalization of external AI tool utilization and internal AI capability. The external AI tools examined mainly refer to platform-based and externally provided AI applications commonly used by small exporting merchants in Yiwu, such as AI translation, product content generation, customer communication, and online promotion support. While this bounded operationalization improves construct clarity and respondent interpretability, the findings should be interpreted within the specific context of Yiwu’s export ecosystem. Accordingly, the negative interaction observed in this study should not be interpreted as evidence that external AI systems and internal AI capability are universally substitutive. Rather, the findings suggest that partial substitution may emerge when standardized external AI tools and internally developed AI capability overlap in supporting routine export-related activities. Future research could examine broader categories of AI systems and develop more generalizable measurement frameworks to further investigate the conditions under which different AI-related resources function as complements or substitutes. In addition, future studies could examine firms across different countries, industries, and digital-platform ecosystems to assess the external validity of the partial substitution effect. Second, this study focuses exclusively on small and micro-enterprises engaged in export activities. Whether similar patterns emerge among larger firms, non-exporting firms, or organizations at different stages of digital transformation remains an open question. Extending the analysis to diverse organizational contexts would help clarify the scope conditions of the proposed framework. Third, although the reliability tests indicate satisfactory internal consistency, the key variables are measured using self-reported survey data, which may raise concerns regarding common method bias or social desirability effects. In addition, the cross-sectional design does not allow us to fully rule out reverse causality. Firms with stronger export performance may have more resources and incentives to adopt external AI tools or develop internal AI capability. Future studies could incorporate objective performance indicators, platform transaction records, financial data, longitudinal designs, panel data, or instrumental-variable approaches to strengthen causal inference and better identify the direction of causality. Longitudinal research would also be particularly useful for examining the dynamic relationship between external AI tool utilization and internal AI capability. While repeated use of external AI tools may facilitate learning-by-using and contribute to internal capability accumulation, excessive reliance on standardized external tools may also reduce firms’ incentives to develop internal AI capability. Future research is therefore needed to clarify whether external AI tool utilization strengthens, weakens, or reshapes internal AI capability over time. 7. Conclusions In conclusion, this study shows that both external AI tool utilization and internal AI capability contribute positively to export performance among exporting small and micro-enterprises. However, their significant negative interaction indicates that these two AI-related resources do not always generate purely additive benefits. Instead, when one resource is already highly developed, the marginal contribution of the other may decline, suggesting partial substitution and diminishing marginal returns. These findings provide useful implications for both researchers and practitioners. For researchers, the study highlights the need to examine not only the independent effects of AI-related resources, but also how different digital resources interact under resource-constrained conditions. For practitioners, especially owner-managers of exporting small and micro-enterprises, the results suggest that AI-related investment decisions should be aligned with firms’ existing resource conditions and operational needs. Rather than assuming that greater investment in both internal AI capability and external AI tools will always improve performance, small firms should carefully balance internal capability development and external tool adoption. Such strategic alignment can help reduce redundant digital investment and support more sustainable export performance improvement. Author Contributions Conceptualization, M.G. and C.J.; methodology, M.G. and C.J.; formal analysis, M.G.; investigation, M.G.; data curation, M.G.; writing—original draft preparation, M.G.; writing—review and editing, M.G. and C.J.; visualization, M.G.; supervision, C.J.; project administration, C.J. All authors have read and agreed to the published version of the manuscript. Funding This research received no external funding. Institutional Review Board Statement This study is waived for ethical review according to the Bioethics and Safety Act in Korea and institutional guidelines governing social science research. Therefore, Ethics Committee approval was not required for this study. All procedures performed in this research were conducted in accordance with the ethical standards of research. Informed Consent Statement Verbal informed consent was obtained from the participants. Verbal consent was obtained rather than written because the data used in this study were collected through an anonymous survey of firms. The survey targeted firms rather than individual human subjects, and no personally identifiable or sensitive personal information was collected. Before completing the survey, respondents were informed about the purpose of the research and that their participation was entirely voluntary. By choosing to complete the questionnaire, respondents provided their informed consent to participate in the study. Data Availability Statement The data presented in this study are available from the corresponding author upon reasonable request. Conflicts of Interest The authors declare no conflict of interest. References Jakubik, A.; Rotunno, L.; Saini, A. Foresee the Unseen: Evaluating the Impact of Artificial Intelligence on International Trade. J. Policy Model. 2025, 47, 842–861. [] [ CrossRef] Ozturk, O. The Impact of AI on International Trade: Opportunities and Challenges. Economies 2024, 12, 298. [] [ CrossRef] Kumar, S.; Vandana; Kumar, V.; Chatterjee, S.; Mariani, M.; De Massis, A. The Role of Artificial Intelligence Capabilities in Enhancing Export Performance: A Study of Ambidexterity and Dynamic Capabilities. Int. Mark. Rev. 2025, 42, 698–714. [] [ CrossRef] Jean, R.-J.; Kim, D.; Cavusgil, S.T.; Chen, C. Determinants of Chinese Exporters’ Online De-Internationalization. Manag. Organ. Rev. 2025, 21, 1110–1130. [] [ CrossRef] Liu, J.; Qin, C.; Chu, X. Development of Corporate Artificial Intelligence and the Quality of Export Products. Financ. Res. Lett. 2025, 78, 107217. [] [ CrossRef] Dong, Y.; He, X.; Blut, M. How and When Does Digitalization Influence Export Performance? A Meta-Analysis of Its Consequences and Contingencies. Int. Mark. Rev. 2024, 41, 1388–1413. [] [ CrossRef] Doan, T.; Luong, D. Effects of Digital Capability on Digital Export: International Evidence. Econ. Bus. Lett. 2025, 14, 166–176. [] [ CrossRef] Li, J.; Chen, L.; Yi, J.; Mao, J.; Liao, J. Ecosystem-Specific Advantages in International Digital Commerce. J. Int. Bus. Stud. 2019, 50, 1448–1463. [] [ CrossRef] Dung Ngo, V.; Leonidou, L.C.; Janssen, F.; Christodoulides, P. Export-Specific Investments, Competitive Advantage, and Performance in Vietnamese SMEs: The Moderating Role of Domestic Market Conditions. J. Bus. Res. 2024, 170, 114315. [] [ CrossRef] Qu, C.; Kim, E. Artificial-Intelligence-Enabled Innovation Ecosystems: A Novel Triple-Layer Framework for Micro, Small, and Medium-Sized Enterprises in the Chinese Apparel-Manufacturing Industry. Sustainability 2025, 17, 5019. [] [ CrossRef] Chen, X.; Wu, Y.; Long, Y. Does Artificial Intelligence Promote Sustainable Growth of Exporting Firms? Sustainability 2025, 17, 7273. [] [ CrossRef] Hasan, R.; Ojala, A. Managing Artificial Intelligence in International Business: Toward a Research Agenda on Sustainable Production and Consumption. Thunderbird Int. Bus. Rev. 2024, 66, 151–170. [] [ CrossRef] Kumar, S.; Kumar, V.; Chaudhuri, R.; Chatterjee, S.; Vrontis, D. AI Capability and Environmental Sustainability Performance: Moderating Role of Green Knowledge Management. Technol. Soc. 2025, 81, 102870. [] [ CrossRef] Han, S.; Zhang, D.; Zhang, H.; Lin, S. Artificial Intelligence Technology, Organizational Learning Capability, and Corporate Innovation Performance: Evidence from Chinese Specialized, Refined, Unique, and Innovative Enterprises. Sustainability 2025, 17, 2510. [] [ CrossRef] Lins, S.; Pandl, K.D.; Teigeler, H.; Thiebes, S.; Bayer, C.; Sunyaev, A. Artificial Intelligence as a Service. Bus. Inf. Syst. Eng. 2021, 63, 441–456. [] [ CrossRef] Nambisan, S.; Zahra, S.A.; Luo, Y. Global Platforms and Ecosystems: Implications for International Business Theories. J. Int. Bus. Stud. 2019, 50, 1464–1486. [] [ CrossRef] Mikalef, P.; Gupta, M. Artificial Intelligence Capability: Conceptualization, Measurement Calibration, and Empirical Study on Its Impact on Organizational Creativity and Firm Performance. Inf. Manag. 2021, 58, 103434. [] [ CrossRef] Cohen, W.; Levinthal, D. Absorptive Capacity: A New Perspective on Learning and Innovation. Adm. Sci. Q. 1990, 35, 128–152. [] [ CrossRef] Li, R.; Wang, Q.; Cheong, K.C. From Obscurity to Global Prominence—Yiwu’s Emergence as an International Trade Hub. Cities 2016, 53, 8–17. [] [ CrossRef] Qian, L.; Lu, P.; Wen, M. Refashioning “the World’s Capital of Small Commodities”: Yiwu’s Internationalization and Digitalization. Cities 2024, 148, 104885. [] [ CrossRef] Shou, X.; Shi, Q.; Zhang, X. The Adaptation and Transformation of Yiwu’s Foreign Trade Enterprises amid Major Changes Unseen in a Century (2001–2021). Transnatl. Corp. Rev. 2024, 16, 200080. [] [ CrossRef] Liu, W.; Si, S. Disruptive Innovation in the Context of Retailing: Digital Trends and the Internationalization of the Yiwu Commodity Market. Sustainability 2022, 14, 7559. [] [ CrossRef] Añón Higón, D.; Bonvin, D. Digitalization and Trade Participation of SMEs. Small Bus. Econ. 2024, 62, 857–877. [] [ CrossRef] Cao, T.L.; Hsu, J. Digitalization and Country Distance in International Trade: An Empirical Analysis of European Countries. Telecommun. Policy 2025, 49, 102877. [] [ CrossRef] Du, X.; Huang, J. The Influence of Digital Capabilities on the Export Performance of SMEs: Evidence from China. Asia Pac. Bus. Rev. 2025, 1–26. [] [ CrossRef] Luu, T.D. Digital Transformation and Export Performance: A Process Mechanism of Firm Digital Capabilities. Bus. Process Manag. J. 2023, 29, 1436–1465. [] [ CrossRef] Oh, S.; Hwang, S. How Does Digital Capability Translate into Export Performance?: The Critical Mediating Role of GVC Upgrading. Asia Glob. Econ. 2025, 5, 100116. [] [ CrossRef] Chishty, S.K.; Sayari, S.; Mohamed, A.H.; Mallick, M.F.; Khan, N.; Inkesar, A. The Utilisation of Artificial Intelligence in the Export Performance of MNCs: The Role of Cultural Distance. Adm. Sci. 2025, 15, 160. [] [ CrossRef] Brynjolfsson, E.; Hui, X.; Liu, M. Does Machine Translation Affect International Trade? Evidence from a Large Digital Platform. Manag. Sci. 2019, 65, 5449–5460. [] [ CrossRef] Dai, J.; Mao, X.; Wu, P.; Zhou, H.; Cao, L. Revolutionizing Cross-Border e-Commerce: A Deep Dive into AI and Big Data-Driven Innovations for the Straw Hat Industry. PLoS ONE 2024, 19, e0305639. [] [ CrossRef] Menzies, J.; Sabert, B.; Hassan, R.; Mensah, P.K. Artificial Intelligence for International Business: Its Use, Challenges, and Suggestions for Future Research and Practice. Thunderbird Int. Bus. Rev. 2024, 66, 185–200. [] [ CrossRef] Ahmad, I. The Strategic Role of Artificial Intelligence in Overcoming Intercultural Barriers for SME Internationalization: A Systematic Literature Review. J. Intercult. Commun. 2025, 25, 148–163. [] [ CrossRef] Yordanova, D.; Dana, L.-P.; Manolova, T.S.; Pergelova, A. Digital Technologies and the Internationalization of Small and Medium-Sized Enterprises. Sustainability 2024, 16, 2660. [] [ CrossRef] Cassia, F.; Magno, F. Cross-Border e-Commerce as a Foreign Market Entry Mode among SMEs: The Relationship between Export Capabilities and Performance. Rev. Int. Bus. Strategy 2022, 32, 267–283. [] [ CrossRef] Barney, J. Firm Resources and Sustained Competitive Advantage. J. Manag. 1991, 17, 99–120. [] [ CrossRef] Chen, D.; Esperança, J.P.; Wang, S. The Impact of Artificial Intelligence on Firm Performance: An Application of the Resource-Based View to e-Commerce Firms. Front. Psychol. 2022, 13, 884830. [] [ CrossRef] Wang, J.; Huang, Q. The Impact of Digital Transformation on the Export Technology Complexity of Manufacturing Enterprises: Based on Empirical Evidence from China. Sustainability 2025, 17, 2596. [] [ CrossRef] Xu, Q.; Li, X.; Guo, F. Digital Transformation and Environmental Performance: Evidence from Chinese Resource-Based Enterprises. Corp. Soc. Responsib. Environ. Manag. 2023, 30, 1816–1840. [] [ CrossRef] Fosso Wamba, S.; Queiroz, M.M.; Pappas, I.O.; Sullivan, Y. Artificial Intelligence Capability and Firm Performance: A Sustainable Development Perspective by the Mediating Role of Data-Driven Culture. Inf. Syst. Front. 2024, 26, 2189–2203. [] [ CrossRef] Falentina, A.T.; Resosudarmo, B.P.; Darmawan, D.; Sulistyaningrum, E. Digitalisation and the Performance of Micro and Small Enterprises in Yogyakarta, Indonesia. Bull. Indones. Econ. Stud. 2021, 57, 343–369. [] [ CrossRef] Salvatierra-Manchego, V.H.; Libaque-Saenz, C.F. Digitalization and Environmental Sustainability as Alternative Strategies to Improve Peruvian MSEs’ Export Performance: A Preliminary Study. Issues Inf. Syst. 2024, 25, 317–331. [] [ CrossRef] Cassiman, B.; Veugelers, R. In Search of Complementarity in Innovation Strategy: Internal R&D and External Knowledge Acquisition. Manag. Sci. 2006, 52, 68–82. [] [ CrossRef] Laursen, K.; Salter, A. Open for Innovation: The Role of Openness in Explaining Innovation Performance among U.K. Manufacturing Firms. Strateg. Manag. J. 2006, 27, 131–150. [] [ CrossRef] Berchicci, L. Towards an Open R&D System: Internal R&D Investment, External Knowledge Acquisition and Innovative Performance. Res. Policy 2013, 42, 117–127. [] [ CrossRef] Ocasio, W. Towards an Attention-Based View of the Firm. Strateg. Manag. J. 1997, 18, 187–206. [] [ CrossRef] Sirmon, D.; Hitt, M.; Ireland, R. Managing Firm Resources in Dynamic Environments to Create Value: Looking Inside the Black Box. Acad. Manag. Rev. 2007, 32, 273–292. [] [ CrossRef] Hambrick, D.C.; Mason, P.A. Upper Echelons: The Organization as a Reflection of Its Top Managers. Acad. Manag. Rev. 1984, 9, 193–206. [] [ CrossRef] [ PubMed] Kumar, N.; Stern, L.W.; Anderson, J.C. Conducting Interorganizational Research Using Key Informants. Acad. Manag. J. 1993, 36, 1633–1651. [] [ CrossRef] Bellandi, M.; Lombardi, S. Specialized Markets and Chinese Industrial Clusters: The Experience of Zhejiang Province. China Econ. Rev. 2012, 23, 626–638. [] [ CrossRef] Sinkovics, R.R.; Kurt, Y.; Sinkovics, N. The Effect of Matching on Perceived Export Barriers and Performance in an Era of Globalization Discontents: Empirical Evidence from UK SMEs. Int. Bus. Rev. 2018, 27, 1065–1079. [] [ CrossRef] Lin, H.F.; Lin, S.M. Determinants of E-Business Diffusion: A Test of the Technology Diffusion Perspective. Technovation 2008, 28, 135–145. [] [ CrossRef] Zhu, K.; Kraemer, K.L. Post-Adoption Variations in Usage and Value of E-Business by Organizations: Cross-Country Evidence from the Retail Industry. Inf. Syst. Res. 2005, 16, 61–84. [] [ CrossRef] Zhang, C.; Bai, T.; Zhou, A.J.; Zhou, S.S. Digital Platforms, Internal Digitalization, and Internationalization of SMEs. Long. Range Plan. 2025, 58, 102588. [] [ CrossRef] Xie, X.; Han, Y.; Anderson, A.; Ribeiro-Navarrete, S. Digital Platforms and SMEs’ Business Model Innovation: Exploring the Mediating Mechanisms of Capability Reconfiguration. Int. J. Inf. Manag. 2022, 65, 102513. [] [ CrossRef] Hagedoorn, J.; Wang, N. Is There Complementarity or Substitutability between Internal and External R&D Strategies? Res. Policy 2012, 41, 1072–1083. [] [ CrossRef] Delgado-Verde, M.; Martín-de Castro, G.; Cruz-González, J.; Navas-López, J.E. Complements or Substitutes? The Contingent Role of Corporate Reputation on the Interplay between Internal R&D and External Knowledge Sourcing. Eur. Manag. J. 2021, 39, 70–83. [] [ CrossRef] Figure 1. Research model. Figure 1. Research model. Figure 2. Interaction effects of external AI tool utilization and internal AI capability on export performance. Figure 2. Interaction effects of external AI tool utilization and internal AI capability on export performance. Table 1. Sample characteristics (N = 475). Table 1. Sample characteristics (N = 475). Firm Characteristics Firm age (years) % Firm size (number of employees) % 1–3 11.37 1 2.32 4–6 41.89 2–5 7.16 7–10 28.00 6–10 15.16 10+ 18.74 11–50 41.05 51–200 26.11 200+ 8.21 Family owned% Production facility % Yes 73.47 Yes 61.68 No 26.53 No 38.32 Main business model% Export-only 40.84 Export-and-domestic 59.16 Respondent Characteristics Gender % Education level % Female 45.05 Junior high or below 5.89 Male 54.95 High school 13.05 College 20.84 Bachelor 45.47 Master or above 14.74 Table 2. Descriptive statistics and correlations. Table 2. Descriptive statistics and correlations. Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (1) Export performance 1.000 (2) External AI tool utilization 0.388 * 1.000 (3) Internal AI capability 0.439 * 0.431 * 1.000 (4) Firm age −0.075 −0.056 −0.041 1.000 (5) Firm size −0.034 −0.030 −0.016 0.043 1.000 (6) Family-owned −0.019 −0.030 −0.001 −0.008 0.058 1.000 (7) Production facility 0.070 0.074 0.149 * −0.009 −0.004 0.125 * 1.000 (8) Main business model −0.018 −0.072 −0.080 0.117 * 0.032 0.072 0.100 * 1.000 (9) Education 0.062 0.088 0.010 −0.003 −0.089 −0.001 0.014 0.028 1.000 (10) Gender −0.023 −0.030 −0.142 * −0.008 −0.059 −0.002 −0.035 0.022 0.036 1.000 Mean 3.726 3.614 3.689 0.533 0.657 0.735 0.617 0.408 0.602 0.451 Std. Dev. 0.927 0.966 0.916 0.499 0.475 0.442 0.487 0.492 0.490 0.498 Note: * p < 0.05. Table 3. Regression results for AI tool utilization, AI capability, and export performance (n = 475). Table 3. Regression results for AI tool utilization, AI capability, and export performance (n = 475). Variable (1) (2) (3) (4) (5) Firm age −0.134 −0.101 −0.109 −0.094 −0.109 (0.086) (0.080) (0.077) (0.075) (0.075) Firm size −0.048 −0.036 −0.033 −0.029 −0.036 (0.090) (0.084) (0.081) (0.079) (0.078) Family owned 0.055 0.028 0.043 0.029 0.042 (0.097) (0.090) (0.088) (0.085) (0.085) Production facility −0.138 −0.077 −0.008 0.000 −0.009 (0.089) (0.082) (0.081) (0.079) (0.078) Main business model −0.029 0.024 0.044 0.060 0.058 (0.088) (0.082) (0.079) (0.077) (0.077) Education 0.113 0.049 0.101 0.064 0.041 (0.087) (0.081) (0.079) (0.077) (0.077) Gender −0.045 −0.023 0.068 0.056 0.055 (0.086) (0.079) (0.078) (0.076) (0.075) External AI tool utilization 0.364 *** 0.228 *** 0.198 *** (0.041) (0.043) (0.044) Internal AI capability 0.447 *** 0.345 *** 0.286 *** (0.043) (0.046) (0.050) External AI tool utilization × Internal AI capability −0.107 *** (0.036) Constant 3.831 *** 3.797 *** 3.688 *** 3.699 *** 3.768 *** (0.121) (0.112) (0.110) (0.107) (0.109) N 475 475 475 475 475 R-squared 0.016 0.157 0.202 0.247 0.262 Adj. R-squared 0.002 0.142 0.188 0.233 0.246 F-statistic 1.11 10.84 14.75 16.99 16.46 Root MSE 0.926 0.858 0.835 0.812 0.805 Standard errors in parentheses.*** p < 0.01. Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. © 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license. Share and Cite MDPI and ACS Style Gu, M.; Jin, C. Artificial Intelligence and Export Performance in Small and Micro-Enterprises: The Roles of Internal Capability and External Tools. Sustainability 2026, 18, 5846. https://doi.org/10.3390/su18125846 AMA Style Gu M, Jin C. Artificial Intelligence and Export Performance in Small and Micro-Enterprises: The Roles of Internal Capability and External Tools. Sustainability. 2026; 18(12):5846. https://doi.org/10.3390/su18125846 Chicago/Turabian Style Gu, Mengyang, and Chuyue Jin. 2026. "Artificial Intelligence and Export Performance in Small and Micro-Enterprises: The Roles of Internal Capability and External Tools" Sustainability 18, no. 12: 5846. https://doi.org/10.3390/su18125846 APA Style Gu, M., & Jin, C. (2026). Artificial Intelligence and Export Performance in Small and Micro-Enterprises: The Roles of Internal Capability and External Tools. Sustainability, 18(12), 5846. https://doi.org/10.3390/su18125846 Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details . Article Metrics Article metric data becomes available approximately 24 hours after publication online.
