The Hiddden Demands of Artificial Intelligence

Abstract

Global AI-led data center growth is driving localized constraints across energy systems and water resources. This project develops a geospatial decision-support framework (AI Risk & Opportunity Index) to assess site-level risk and opportunity in the United States, reframing AI infrastructure as a location-dependent physical system. Leveraging a GIS-based modeling approach, the methodology applies spatial overlays, normalization, and weighted scoring to develop a rounded view of site-level risk including grid stress (emission intensity, energy mix , load capacity, congestion, and price), water stress (basin stress, availability, and quality), resource intensity (land use, ecosystem sensitivity), and overall regulatory and reputational risk. The framework integrates site-level scores with company transition readiness to enable geo-spatially intelligent decisions for stakeholders.