From AI-enabled materials informatics to sustainable innovation in Asian global value chains: the bridging role of inter-organisational collaboration

The twin transition of digitalisation and sustainability requires manufacturers in Asian global value chains (GVCs) to convert internal artificial intelligence (AI) capability into outcomes that depend on cross-boundary coordination. Drawing on the Dynamic Capabilities View and GVC governance research, this study examines how AI-enabled materials informatics (MI) capability relates to sustainable innovation performance (SIP) through inter-organisational collaboration. Survey data from 205 senior technology executives in Taiwan's high-technology manufacturing sector were analysed using a complementary quantitative dual-analytic strategy combining PLS-SEM and fsQCA. PLS-SEM shows positive associations between AI-enabled MI capability and collaboration capability (β = 0.612, p < .001), collaboration capability and SIP (β = 0.485, p < .001) and AI-enabled MI capability and SIP (β = 0.215, p = .002). The specific indirect association is significant (β = 0.297, p < .001), and sensitivity analysis excluding SIP_2 yields materially unchanged results. fsQCA identifies three sufficient configurations for high SIP, including an Eco-Collaborative pathway combining weak internal AI capability with strong data-governance and regulatory-sensing practices. The findings highlight complementary net-association and configurational perspectives on sustainable innovation.