Mississippi River Mayors Pursue Water Compact to Shield Basin From AI Data Centers
Mayors from 10 states along the Mississippi River are pushing for a federally recognized compact to ban the mass exportation of freshwater, citing the massive cooling demands of the rapidly expanding AI data center industry.
Share this story
A single hyperscale artificial intelligence data center can consume 5 million gallons of freshwater every day—the exact equivalent of the daily water usage for a town of 10,000 people. As the infrastructure required to train and run generative models expands, the sheer physical volume of that consumption has transformed AI from a digital abstraction into a localized resource competitor.[2]
That physical reality drove mayors from 10 states along the Mississippi River to convene in Natchez, Mississippi, this week to establish a collective defense of their water supplies. The coalition, representing the Mississippi River Cities and Towns Initiative (MRCTI), announced a coordinated legislative push to ban the mass exportation of freshwater from the river basin to cool AI data centers.[1][3]
The scale of the incoming infrastructure is unprecedented. Construction of data centers in the United States has reached a record $4.4 billion per month, with hyperscale AI spending projected to near $700 billion this year. Illinois alone currently houses 139 active data centers with another 100 in development, making it the most concentrated hub along the river. According to Blytheville, Arkansas, Mayor Melisa Logan, co-chair of the MRCTI, more than 75 percent of the coalition's nearly 200 member cities have already been approached by companies seeking to host AI facilities.[3][4]
To protect the 2,000-mile river corridor, the mayors are modeling their strategy on the 2008 Great Lakes Compact, which strictly prohibits water from being piped out of the Great Lakes basin. The Mississippi River effort relies on a state-by-state legislative domino effect, beginning with a 2025 Missouri law that requires a permit for anyone withdrawing water and exporting it more than 30 miles outside the state.[1][2]
Lawmakers in Mississippi, Louisiana, and Minnesota have now committed to introducing identical bills. If six of the ten river states pass the export ban, the coalition plans to elevate the agreement to Congress for formal federal recognition as the Mississippi River Compact.[2][5]
The push for a binding compact illustrates a fundamental shift in how local governments view the tech industry's expansion. While previous generations of digital infrastructure primarily strained electrical grids, the cooling requirements of high-density AI clusters present a direct physical threat to municipal water security, particularly as droughts and extreme heat events become more frequent.[2][3]
"Water is not simply an environmental resource," Natchez Mayor Dan Gibson noted during the summit, emphasizing that the river drives the economy, culture, and basic survival of hundreds of communities. In Louisiana alone, water-intensive industries like fisheries and waterborne commerce account for nearly 80 percent of the state's economy.[1][3]
The MRCTI has also released a new evaluation framework, developed alongside the Sierra Club, to help local officials weigh the fiscal returns of data center proposals against their resource usage and public cost. As the AI arms race accelerates, the Mississippi River Compact represents the first major regional attempt to ensure that the physical footprint of the next industrial revolution does not drain the natural resources of the communities hosting it.[4][5]
Viewpoints in depth
Municipal Water Defenders
Local leaders and environmental groups prioritizing the physical security of the river basin over unrestricted tech expansion.
For the mayors of the MRCTI and allied groups like the Sierra Club, the abstraction of 'the cloud' has been replaced by the physical reality of cooling towers and evaporation rates. They argue that without a binding interstate compact, the immense capital behind AI development will simply pit river towns against one another, extracting millions of gallons of freshwater daily. By establishing a unified front modeled on the Great Lakes Compact, they aim to force hyperscale developers to adapt their cooling technologies rather than draining municipal supplies during periods of increasing drought.
AI Infrastructure Developers
Technology firms racing to secure the power and water necessary to train next-generation artificial intelligence models.
The technology sector views the rapid deployment of data centers as a critical national security and economic imperative, with monthly construction spending now exceeding $4.4 billion. While acknowledging the resource intensity of these facilities, developers argue that they are investing heavily in closed-loop cooling systems and acoustic mitigation. For these firms, securing locations along major waterways like the Mississippi is essential to maintaining the United States' competitive edge in the global AI arms race, making local moratoriums a bottleneck to innovation.
Key points
- Mayors from 10 states along the Mississippi River are drafting a compact to protect the basin's water from AI data center consumption.
- A single hyperscale data center can use 5 million gallons of water daily, equivalent to a town of 10,000 people.
- The coalition is modeling its effort on the Great Lakes Compact, aiming for a federal ban on mass water exportation.
- More than 75 percent of the coalition's member cities have already been approached to host AI infrastructure.
How we got here
2008
The Great Lakes Compact is signed into law, establishing a precedent for banning the export of water from a major US basin.
2025
Missouri passes legislation requiring a permit for anyone exporting water more than 30 miles outside the state.
September 2026
Mayors from 10 Mississippi River states convene in Natchez, Mississippi, to draft a unified compact regulating AI data center water usage.
- Municipal Water Defenders
- Local leaders and environmental groups prioritizing the physical security of the river basin over unrestricted tech expansion.
- AI Infrastructure Developers
- Technology firms racing to secure the power and water necessary to train next-generation artificial intelligence models.
- Economic Growth Advocates
- Local economic boards and legislators seeking to balance resource protection with the massive tax revenue data centers promise.
Perspectives this story doesn't cover
- Hyperscale Cloud Providers
- Local Utility Ratepayers
Sources
[1]Circle of BlueMunicipal Water DefendersThe Stream, September 22, 2026: Mississippi River States Launch Effort to Protect Water Resources from AI Data Centers
Read on Circle of Blue →
[2]News From The StatesMunicipal Water DefendersMississippi River mayors call for water conservation protections amid growing data center demand
Read on News From The States →
[3]WBRZMunicipal Water DefendersMayors from states along Mississippi River meet to develop plans for AI infrastructure, water supply security
Read on WBRZ →
[4]Mississippi River Basin Ag & Water DeskEconomic Growth AdvocatesMississippi River mayors help cities navigate rising data center interest
Read on Mississippi River Basin Ag & Water Desk →
[5]DeSoto County NewsAI Infrastructure DevelopersMississippi River mayors seek water conservation amid data center demand
Read on DeSoto County News →
More in Opinion
See all →Automotive Engineering
How Positive Crankcase Ventilation Valves Regulate Blowby Gas Flow Across Shifting Engine Manifold Vacuums
4 sources
Atmospheric Optics
Fried's Atmospheric Coherence Length Limits Ground Telescopes to 20-Centimeter Aperture Resolution Regardless of Mirror Size
4 sources
Fluid Thermodynamics
Urea's 32.5 Percent Eutectic Threshold Prevents Fractional Freezing, Keeping Thawed Diesel Exhaust Fluid Chemically Uniform
4 sources
Semiconductor Physics
Why Opposing Zener and Avalanche Temperature Coefficients Cancel at 5.6 Volts
10 sources
Comments
Every angle. Every day.
Get Opinion stories with full source coverage and perspective breakdowns, free every day.




