WATer and the Sustainability of Artificial Intelligence: political economy, infrastructure and the water-energy nexus

Objective

The increase in production and scale of AI infrastructure over recent years has been accompanied by an equally dramatic rise in the use and demand for water as an increasingly scarce resource. The project interrogates the sustainability of AI by examining its environmental impact, in particular the depletion of water resources, exploring the relation between data centers and the reshaping of waterscapes. Through a political economy approach, such analysis brings together a consideration of the role of the state and public policies, the strategies of companies, and the response of social movements, as the construction of data centers raises crucial questions of water justice and water grabbing, especially in areas already facing water shortages. Methodologically, the analysis rests on multi-sited fieldwork, comprising of semi-structured interviews, events participation and documentary analysis, to conduct an incorporated comparison of the US Southwest (Arizona and Texas) and the Atlantic Archipelago (UK and Ireland), with a focus on energy-water entanglements and their political implications. The cases provide a variety of landscapes in which data centers are located, providing insights on the impact not only on environmental resources, but importantly on social relations. The findings address the pressing societal question of reconciling AI technological development with environmental sustainability, particularly water resources, addressing several Sustainable Development Goals. Building on my specialisation in political economy and water governance, through WATSAI I expand it and take it forward by applying to a contemporary societal issue that is AI infrastructural development. The institutions involved bring complementary expertise to the project: water governance (Prof. Menga, University of Bergamo, main host); political economy (Prof. Akbulut, Concordia University, outgoing phase), AI politics (Prof. Bailey, University of Birmingham, secondment).