Beyond “If”: How, Where, and Why Artificial Intelligence is Reconfiguring GVCs

The proliferation of artificial intelligence (AI) is poised to reconfigure the global operations of multinational firms, yet its precise impact on global value chain (GVC) dynamics remains theoretically and empirically underexplored. This study addresses this gap by applying a novel methodology to measure and categorize firm-level AI adoption. Using a deep learning approach (e.g., BERT model) on the text of US firms’ 10-K filings, we construct multi-dimensional indicators for the intensity, diversity, and strategic type of AI adoption. Challenging the traditional view that AI adoption enhances operational stickiness in existing production locations, our research offers a different view on the contingent relationship between AI adoption and GVC relocation and tests it empirically. We argue that the impact of AI is not uniform, but that its effect is contingent on the type of AI adopted, the firm’s strategy, its industry context, and its position within the supply chain network. This study contributes to the literature by providing an evidence-based framework for understanding which types of AI adoption, and under what strategic conditions, are reshaping the global corporate landscape.