Your cloud dries their river.
Literal meaning: A structural relationship in which digital infrastructure extracts resources, labour, and data from lower-income countries while the economic and technological benefits flow primarily to wealthier ones. In the AI energy context, the term focuses specifically on the energy and environmental dimension of that extraction.
Origin: The term draws on postcolonial theory and dependency economics. Scholars Nick Couldry and Ulises Mejias developed the concept in The Costs of Connection (2019). In the AI energy debate, it gained sharper specificity as researchers documented the geographic concentration of data centre construction in regions with regulatory gaps and resource abundance.
The pattern in which AI infrastructure consumes local resources — energy, water, land — in lower-income regions while the benefits flow elsewhere.
The Appeal: Data centres bring jobs, investment, and infrastructure to regions that may lack them. Host governments actively court these investments. The framing of mutual benefit is not always empty — employment, grid upgrades, and local supplier chains are real.
The Friction: The structural asymmetry persists beneath the mutual-benefit rhetoric. Externalized Costs concentrate in the host region. Sacrifice Zones are selected in part because political capacity to resist is lower. Data about local populations is often part of the extraction: training datasets draw from global commons while serving predominantly wealthy users. Couldry and Mejias named the pattern directly: data extraction follows the routes of historical resource extraction.
Why This Matters: Digital colonialism asks where the infrastructure lives and who benefits. Once you know the term, “cloud” stops being a neutral metaphor. Clouds are in specific places, drawing on specific resources, chosen for reasons that have everything to do with power.
Related terms: AI Energy Consumption · Sacrifice Zones · Externalized Costs · Data Center Water Consumption · Greenwashing · Training Run
Read more:
- The Costs of Connection — Couldry, N., & Mejias, U. A. (2019). Stanford University Press
- Making AI Less Thirsty — Li, P. et al. (2023). arXiv