AI Needs a Grange
The opinion article argues that local communities should retain authority over data center siting and other AI-related decisions because they bear the projects' impacts on water, electricity, land, jobs and local character. It calls for greater transparency from developers and government officials, while reserving federal action for problems that cross local or state boundaries.
Eli Lehrer of the R Street Institute and Francesca de Quesada Covey of the Knight Foundation use the historical Grange movement to argue that AI governance should preserve local decision-making. They distinguish between globally connected AI models, which may require national or international rules, and physical data centers, whose effects on land, water, power, noise and tax bases are primarily experienced by host communities.
The authors contend that local governments should be free to approve or reject data centers, provided residents receive accurate information about water use, electricity demand, jobs and subsidies. They cite a 2026 data center deal negotiated under nondisclosure agreements in Louisiana's Rapides Parish and Pennsylvania Governor Josh Shapiro's subsequent decision to bar agencies under his control from using such agreements for data center projects. The article favors disclosure, open markets and accountability over uniform state mandates or federal preemption, except where impacts clearly cross jurisdictional boundaries.