The Infrastructure of Intelligence: Navigating Siting Conflicts, Community Equity, and Grid Resilience in the AI Era

Session Submission Type: Panel

Abstract

The United States is currently undergoing a hyperscale data center construction boom, catalyzed by the rapid expansion of generative Artificial Intelligence (AI). These facilities represent a new class of high-density load infrastructure that consumes energy equivalent to mid-sized cities and requires millions of gallons of water daily. While critical for technological progress, the speed and scale of this deployment have outpaced existing regulatory frameworks, leading to intensified conflicts over land use, energy justice, and grid reliability. 

This panel advances a multi-scalar analysis of AI infrastructure governance, tracking the "lifecycle" of data center siting from local zoning negotiations and state regulatory policy design frameworks to public perception and empirical grid outcomes. The session comprises four complementary papers that address the emergence of data center governance through institutional, distributional, stakeholder, and power-capacity lenses: 
1) Dong and Hsueh (ASU) establish the national landscape, using a large-language-model-fine-tuned BERTopic analysis to reveal how local "package deals" have evolved from simple zoning standards to complex community benefit agreements.
2) Paul (MIT) scales this analysis to the state level, providing a comparative study of Louisiana and Minnesota. This research introduces a "Multifactor Integrated Regulatory Framework" (MIRF) to address systematic gaps in environmental justice and procedural access for marginalized communities.
3) Zheng and Wang (Iowa State) present results from an original conjoint survey experiment of over 1,000 residents. This paper identifies the specific attributes—such as utility bill impacts and residential proximity—that determine the “social license to operate” and tests how “technological progress” versus “regional competition” narratives shift public support.
4) Ye et al. (Maryland/Ohio State) conclude the session with a nationally representative empirical analysis of grid impacts. Using two-way fixed-effects modeling across 14 states, they test whether data center expansion degrades or—through infrastructure reinforcement—improves community-level power resilience. 

Collectively, this panel provides a comprehensive evaluation of the distributional consequences of the AI boom. By bringing together scholars from four distinct geographic regions (Southwest, Northeast, Midwest, and Mid-Atlantic) and various career stages, the session offers a theory-driven and rigorous assessment of how policy frameworks can be redesigned to ensure that the infrastructure of the future is both resilient and equitable.

Policy Area

• Natural Resource, Energy, and Environmental Policy

Chair

• Lily Hsueh, Arizona State University

Discussant

• Tyler A Scott, University of California, Davis

Organizer

• Lily Hsueh, Arizona State University

Individual Presentations

• The Evolution of Data Center Siting Governance in the U.S.: Zoning, Compensation, and Collaboration - Presenting Author: Ximing Dong, Arizona State University

• Governing Data Centers in the Public Interest: State Regulatory Frameworks and Environmental Justice - Presenting Author: Sanjana Paul, Massachusetts Institute of Technology

• How Citizens Perceive Local AI Data Center Development: Economic Benefits, Environmental Tradeoffs, and Policy Framing - Presenting Author: Guimin Zheng, Iowa State University

• Data center expansion and community power resilience in the USA - Presenting Author: Xiaofeng Ye, University of Maryland, College Park