Sustainable and Ethical AI Infrastructure

The rapid expansion of AI data centers is creating unprecedented demand for electric power and water resources. To meet peak power demand, fossil fuel power plants are already running close to full capacity and more are being deployed to address increasing data center demand. However, these create major greenhouse gas emissions. Data center power demand is projected to soar further, raising concerns about grid strain and costs. AI data centers also place growing pressure on water resources for cooling, raising additional concerns regarding water conservation and sustainable infrastructure planning.

This initiative will integrate cross-disciplinary expertise to not only provide input to help guide the State of Maryland’s policy for AI and quantum data center deployment in the coming years, but to chart a national and global path toward sustainable and ethical AI infrastructure. The team will assess four core areas of impact for Maryland arising from the rapid growth in AI data centers: (1) electricity affordability and how policies impact the burden of increased generation costs from new data center loads, (2) the ability of the state to effectively balance economic growth incentives with climate and environmental goals, (3) the critical national security risks posed by threats to large data centers and associated electricity infrastructure, and (4) air quality impacted by emissions from out-of-state electricity producers, such as coal-fired generators. The team will examine technology and policy solutions—both near-term (1-3 years) and longer-term (3-10 years)—that can help address these core areas of impact, while at the same time generating economic growth and business development opportunities. The team will explore emerging technologies such as solid oxide fuel cells, sodium-ion batteries, next-generation small modular reactors, power electronics, and back-up fuel production.

This initiative will integrate engineering analysis, economic modeling, geographic information system (GIS) and machine learning, and policy design to assess how data center infrastructure growth affects grid reliability and environmental costs, community and household energy security, equitable cost-allocation mechanisms, and other areas. The team will share its work with the public and will engage policy makers in the state and beyond to help ensure a more ethical and sustainable infrastructure for AI and quantum data centers.