Industry analysts estimate global investment in AI computing infrastructure could exceed $1 trillion over the next decade as companies race to build capacity for artificial intelligence services, cloud platforms, streaming media and enterprise applications.
States across the country have begun competing to attract these projects with tax incentives, infrastructure support and favorable regulatory policies.
Michigan leaders have increasingly viewed data centers as a potential new pillar of the state’s evolving technology economy.
Proposed AI Data Center Near Saline Highlights Debate
The issue has already surfaced in Southeast Michigan.
A proposed large data center project in Saline Township in Washtenaw County has drawn attention due to its potential scale and links to the expanding artificial intelligence industry.
Local officials have been evaluating plans for a facility that could eventually require hundreds of megawatts of electricity, making it one of the largest technology infrastructure projects in the region.
While supporters say such projects can bring construction jobs and investment, residents have raised concerns about land use, energy consumption and long-term community impact.
Legal challenges and zoning disputes surrounding the project have already highlighted tensions between economic development and local concerns.
Why AI Data Centers Are Suddenly Everywhere
The explosive growth of artificial intelligence is triggering a massive surge in data center construction across the United States.
Key drivers include:
• AI computing demand – Training and running large AI models requires enormous processing power.
• Cloud services growth – Businesses increasingly rely on cloud platforms to run software and store data.
• Streaming and digital services – Video, gaming and financial transactions all depend on large computing networks.
• Technology competition among states – Governments are offering incentives to attract billions in investment.
Modern hyperscale data centers can consume 100 to 500 megawatts of electricity, roughly equivalent to the power used by tens or even hundreds of thousands of homes.
Energy analysts estimate data centers could account for as much as 8 percent of total U.S. electricity demand by the end of the decade, more than double current levels.
Electricity Demand Raising Grid Concerns
The enormous energy consumption of AI infrastructure is one of the primary concerns driving the proposed legislation.
A single hyperscale data center can require as much power as a small city, forcing utilities to consider major upgrades to transmission lines and generation capacity.
Michigan utilities are already preparing for rising electricity demand from:
• electric vehicle manufacturing
• battery plants and advanced manufacturing
• broader electrification of industrial processes
Adding multiple large data centers to that mix could accelerate the need for additional power plants and grid infrastructure.
Lawmakers supporting the moratorium say the state needs a comprehensive plan to manage those demands before approving a wave of new projects.
Communities Question the Economic Trade-Off
Although data centers can involve hundreds of millions or even billions of dollars in investment, they often create relatively few permanent jobs once construction is complete.
Most employment occurs during the building phase.
Once operational, a large facility may employ only 50 to 100 full-time workers.
That has led some communities to question whether the infrastructure demands — particularly electricity and water usage — justify the long-term economic benefits.
Concerns often raised by local officials include:
• heavy electricity demand that could strain regional grids
• water consumption needed to cool massive server farms
• tax incentives offered to technology companies
• large land footprints that can cover dozens or even hundreds of acres
Supporters counter that data centers generate property tax revenue and help position regions as technology hubs capable of attracting additional investment.
Which Michigan Communities Could See Data Centers Next
If Michigan ultimately embraces large-scale data center development, several regions could emerge as potential hubs.
Industry analysts and economic development officials often look for locations with strong power infrastructure, fiber networks and available land.
Areas frequently mentioned as potential candidates include:
Washtenaw County – Already home to major technology companies and research institutions, including the University of Michigan.
Detroit and Wayne County – Existing fiber infrastructure and proximity to international data networks make the region attractive for cloud computing operations.
Grand Rapids region – West Michigan has seen growing technology investment and has access to major power transmission lines.
Lansing and mid-Michigan – The presence of state government and expanding infrastructure could make the region attractive for future projects.
Northern Michigan – Cooler climates can reduce cooling costs for data centers, making some northern locations appealing for developers.
Each of these regions offers a combination of land availability, energy access and fiber connectivity that technology companies typically seek when evaluating new data center sites.
Michigan Competing in the AI Infrastructure Race
Across the United States, states including Virginia, Texas, Ohio and Arizona have emerged as major data center hubs.
Northern Virginia’s “Data Center Alley,” for example, hosts one of the world’s largest concentrations of internet infrastructure.
Michigan policymakers hope the state can capture some of the next wave of AI-driven technology investment as part of a broader strategy to expand beyond its traditional manufacturing base.
But the proposed moratorium highlights a fundamental question facing lawmakers:
How quickly should Michigan expand AI infrastructure — and how much energy and land should be devoted to powering the next generation of computing?
The answer could shape Michigan’s role in the rapidly evolving artificial intelligence economy for decades to come.