Overview:
- Michigan's AI data center pledge includes promises to pay infrastructure costs, add clean energy, protect water resources, and hire local workers.
- A public scorecard would track whether companies actually deliver on commitments by measuring costs to ratepayers, new power generation, water use, and local hiring, says Gleb Tsipursky.
- Transparent reporting would help communities make informed decisions about future data center projects based on evidence rather than competing claims.
Michigan is trying to solve two problems at once. The state wants the investment and computing capacity associated with artificial intelligence, while residents want reliable power, affordable bills, and protection for water and other local resources.
The tension is becoming harder to ignore. A federal appeals court ruled Sept. 11 that the U.S. Department of Energy exceeded its emergency authority when it ordered the J.H. Campbell coal plant in West Olive to remain open after its planned retirement. The plant’s extended operation has become part of a larger argument over reliability, cost, and who should pay when the grid is asked to do more.
Artificial intelligence data centers make that argument more urgent because their electricity demand can be enormous, and arrive faster than traditional planning cycles.
Michigan has already adopted the right principle. In July, Gov. Gretchen Whitmer announced the Michigan Affordability and Responsible Growth Pledge, under which participating AI and data center companies agree to bear the costs directly attributable to their facilities, invest in additional power resources, pursue clean energy, hire Michiganders, and safeguard natural resources.
The next step should be to make those promises measurable in public.
Michigan should create a simple data center scorecard that residents, regulators, local officials, and companies can all read. It should answer four questions.
First, are ordinary customers actually insulated from data center costs? Companies can promise to pay for new generation and grid upgrades, but the public needs an understandable accounting of what was added, what it cost, who paid, and whether any cost shifted onto households or small businesses.
Second, how much new power is truly additive? A company that contracts for clean generation should report how much capacity came online because of the project, when it became available, and how its output compares with the facility’s demand. The point is not to force every project into one technology. It is to distinguish new supply from accounting that merely reallocates existing resources.
Third, what is happening to water? Communities evaluating large facilities should be able to see expected and actual water withdrawals, the cooling approach used, wastewater impacts where relevant, and contingency plans for drought or local supply constraints. These are practical planning questions, not ideological ones.
Fourth, what does the community receive beyond construction activity? Michigan’s pledge includes hiring residents. That should become a measurable workforce commitment: local hiring, apprenticeships, technician training, and the number of workers who move into higher-skill roles over time.
A scorecard would improve the debate because it would shift attention away from competing forecasts and toward results.
Supporters of data centers could point to evidence that a project paid its own infrastructure costs, added power, managed water responsibly, and created durable local jobs. Critics could identify where promises were not kept. Local governments considering the next proposal would have Michigan evidence rather than relying mainly on claims from developers or opponents.
The same logic should apply to environmental safeguards. A rule matters most when people can see whether it is working. If a project commits to a particular water strategy, grid investment, or clean energy addition, the relevant outcome should be reported on a regular schedule in the same place.
This is especially important because AI infrastructure is moving quickly. Communities should not have to choose between embracing a new industry blindly and blocking it reflexively. They need a feedback system that lets them approve projects with clear conditions, observe what happens, and tighten or relax those conditions based on evidence.
Michigan has an opportunity to set that standard. Its affordability pledge already recognizes the core political reality: public support for AI infrastructure will depend on whether residents believe they are being asked to subsidize private growth or absorb risks they did not choose.
A transparent scorecard would turn that principle into something stronger than a promise.
It would make it testable.
Planet Detroit’s Voices column includes opinion pieces from our community of partners and readers. These pieces express the voices of the authors and not necessarily those of the publication.
MORE PLANET DETROIT VOICES
Voices: $5 gas isn’t just a price spike. It’s a warning.
As gas prices reach $5 per gallon across Michigan, the surge highlights America’s persistent oil dependence and its impact on working families — from the pump to the grocery store.
Voices: How clean energy can drive Detroit’s growth
Updating old building stock not only improves energy efficiency and affordability, it’s also an economic engine for the neighborhoods that need it most, says Elevate CEO Anne Evens.
Voices: Michigan PFAS Awareness Week is a call to action
Michigan’s data center expansion could introduce new sources of PFAS contamination to communities already burdened by forever chemicals in their water, soil, and air, says Michigan Green Muslims director.
