Configurations of Buyer–Supplier AI Orientations and Firm Resilience: A Matching Perspective

Purpose Drawing on matching theory, this study examines how configurations of artificial intelligence (AI) capability orientations – automation and smartness – within buyer–supplier dyads influence buyer firm resilience. Design/methodology/approach We employ Bidirectional Encoder Representations from Transformers (BERT), a deep learning approach, to measure firms' AI automation orientation and AI smartness orientation based on annual report disclosures. Using a sample of 3,279 buyer–supplier dyads from Chinese A-share listed firms between 2018 and 2022, we test our hypotheses with ordinary least squares (OLS) regression models. Findings The results show that complementarity-based configurations, specifically buyer automation combined with supplier smartness and buyer smartness combined with supplier automation, significantly enhance buyer resilience. A compatibility-based configuration in which both buyer and supplier emphasise automation also strengthens resilience. In contrast, the matching between buyer smartness and supplier smartness is negatively associated with firm resilience. Originality/value This research distinguishes between AI automation and smartness orientation and demonstrates how their configurations across buyer–supplier dyads generate compatibility, complementarity, or friction. By moving beyond firm-centred and monolithic views of AI, the study offers a novel framework for understanding when dyadic AI alignment enhances resilience and when it undermines it, providing actionable insights for AI-enabled supply chain management in disruption-prone environments.