Send message to Author

Paper short abstract

This paper examines the global value chain of artificial intelligence, tracing its global production. It shows that the Global South—China excepted—contributes labour-intensive activities and primary resources, while (tenuous) value capture remains concentrated elsewhere.

Paper long abstract

This paper analyses the global value chain of artificial intelligence, focusing on the geographical distribution of activities involved in the production, deployment, and use of AI systems. It maps key stages of the AI value chain—including data extraction and labelling, computing infrastructure, model development, and deployment—and examines how value and power are allocated across regions. The analysis shows that the Global South, with the notable exception of China, contributes extensively to AI production through labour-intensive tasks and the supply of primary resources mainly minerals, and data - but also energy and water. These contributions are essential to the functioning of contemporary AI systems, yet they tend to be weakly recognised in the existing value extraction strategies.

The paper further considers the broader economic and social implications of AI. Contrary to dominant narratives of rapid technological transformation, it finds that the impacts of AI remain highly uncertain and uneven. Empirical evidence suggests that AI diffusion has so far been limited, with adoption proceeding at a relatively mild pace in both the Global North and the Global South. Rather than a generalised AI revolution, current patterns point to selective and experimental uptake, concentrated in specific sectors and firms. Taken together, the findings call for a more grounded and globally attentive understanding of AI production and its developmental consequences.