Iris Stewart-Frey

The proposed study will investigate the extent to which the expansion of hyperscale data centers for generative AI affects water scarce and socio-economically vulnerable regions in California and Europe. Areas in both California and Europe have seen some of the most rapid growth of data centers, growing water scarcity is a concern, and they are at the forefront of developing environmental regulation that can serve as a role model for other regions and provide incentives for greater sustainability for a rapidly expanding industry. Using geospatial analysis, a comparative case study approach, and interviews, it will integrate quantitative and qualitative data to examine water use by data centers, specifically for on-site cooling, and its impact on local water availability. The research explores water sources and water resource impacts, data transparency gaps, regulatory responses, and socio-environmental justice concerns. By combining environmental data with policy and through the development of an ethics framework, the project aims to inform sustainable water governance and guide equitable decision-making in the context of AI infrastructure growth. Greater transparency on the water use of hyperscale data centers and their impacts on water supplies could spur competition towards greater water efficiency and not only greater environmental but also economic sustainability.