About Midwest AI Watch

What is Midwest AI Watch?

Midwest AI Watch is a public-interest project that tracks data center development across seven Midwest states: Minnesota, Wisconsin, Illinois, Michigan, Indiana, Ohio, and Iowa. We compile publicly available information into an interactive map so residents, journalists, and policymakers can see what's being built, where, and at what scale.

Why does this exist?

The AI infrastructure boom is transforming communities across the Midwest. Data centers require massive amounts of electricity and water, and their construction reshapes local land use, tax structures, and power grids. These impacts deserve transparent, accessible tracking, not buried in county planning documents or corporate press releases.

We believe informed communities make better decisions. Midwest AI Watch exists to put this information in one place, sourced and cited, so anyone can understand what's happening in their region.

Where we stand

We aren't entirely against AI. Technology will always find ways to evolve and improve over time, and AI is one of, if not the most transformative leaps in technology we've ever experienced. We use the tools ourselves. The infrastructure will live somewhere. It just doesn't have to be in our backyards.

What we are against is the way the siting often goes. Data centers are getting built next to homes, schools, and small towns with little real community input. Local officials approve rezonings and tax abatements on accelerated timelines. Environmental reviews get waived. Water rights get traded for promised jobs that frequently don't materialize at the scale advertised. The benefits (earnings, megawatts, AI capacity) flow to operators and shareholders. The costs fall on the families that have to live with the noise, the lights, the truck traffic, the strain on local water and power.

That asymmetry is the editorial focus of this site. We don't think it has to be that way, and we don't think communities can push back on what they don't know about.

So we track what's actually happening, facility by facility:

• Land: parcel size, current ownership, what the site was zoned for before construction
• Power: planned and current megawatts, utility, interconnection status, transmission upgrades required, on-site generation
• Water: gallons per day, source, cooling type, wastewater destination, water rights acquired
• Proximity: distance from the nearest homes, schools, and hospitals
• Money: tax abatements, TIF districts, PILOT agreements, projected and delivered jobs, average wage
• Approvals and reviews: zoning approval dates, hearings held, environmental review status, air-quality and stormwater permits, lawsuits
• Local response: local officials' stated positions, petitions, opposition groups, supporting groups, organized pushback

We don't decide which projects are worth it. We try to make sure the people who do have the information to make that call honestly: local councils, county boards, planning commissions, and the residents who will live next to them.

How is the data verified?

Each entry comes from public sources: news outlets, county planning documents, government filings, agenda minutes, FOIA requests, and official company announcements. Our pipeline has three layers:

• Automated verification: a Gemini-driven scanner cross-references multiple public sources for each facility weekly, extracts structured fields from articles and filings, and flags entries that need a human look.
• Approved contributors: vetted volunteers submit per-field updates with cited sources (news articles, permit numbers, planning documents, uploaded PDFs). Other contributors can vouch or flag a submission to help prioritize review.
• Admin review: every contributor submission gets read by an administrator, who confirms the citation actually supports the proposed value before it lands in the database. AI-graded relevance checks help speed this up.

The published record is the aggregate of those three streams. Submissions go on the public timeline as facts about the facility, not as bylined posts; individual contributor names aren't shown publicly. We prioritize credible, verifiable reporting over speed.

How can I contribute?

Know about a data center that's not on our map? Submit a report and our team will verify it. You can also suggest updates to existing facilities using the "Suggest an edit" button on any facility's detail card, or apply to become an approved contributor to submit updates directly.

If you're a journalist, researcher, or community organizer who'd like to collaborate, reach out to us at contact@dpdigital.io.