Key points are not available for this paper at this time.
Purpose Generative Artificial Intelligence (Gen-AI) has been rapidly adopted into organizations; however, the process of how technical implementations match human-driven supply chain dynamics is still understudied. Adoption of technology has previously been viewed as a univariate success–failure metric, failing to account for the underlying configurational complexity of implementation. To address this gap, our research uses configurational theory as a foundation to investigate what set of organizational, technical, human, and external conditions is necessary and sufficient to drive Gen-AI implementation as human-driven supply chain facilitators. Design/methodology/approach The research is based on empirical evidence of 137 respondents from manufacturing firms. Utilizing fuzzy-set Qualitative Comparative Analysis (fsQCA), this study finds several equifinal combinations of conditions associated with successful Gen-AI implementation and presents causal asymmetry between high and low implementation conditions. Findings The findings identify “Scalability, Skills and Expertise, Collaboration & Communication, and Industry Trends” as necessary conditions for the successful implementation of Gen-AI. Additionally, seven sufficient configurations exhibit extreme equifinality and causal asymmetry in the demonstration of successful implementation, indicating that skills and collaboration were universal core conditions present in each of the seven configurations. Additionally, on the one end of the spectrum, there are solution sets consisting of enabling technology, assuming that “leadership vision, resource availability, innovation infrastructure, and market competition” will complement technological readiness. On the other hand, human factors focus on the change management aspects, like “organisational culture, user acceptance, and scalability”. The results from each subsample and holdout analysis add further assurances of the predictive validity of configurational analysis in determining how to implement Gen-AI in human-centric supply chain operations. Originality/value This research contributes to supply chain literature by providing a configurational view of Gen-AI implementation. Instead of continuing the debate of “whether implementation works” by conducting additive analyses, our study unlocks and explains equifinal and asymmetrical causal patterns, thereby offering theory-grounded yet managerially applicable pathways on “how different pathways work” for managers to simultaneously leverage technology acceleration and human empowerment in supply chain ecosystems.
Samadhiya et al. (Tue,) studied this question.