A University of Washington student has turned the messy, town-by-town backlash against AI data centers into something anyone can actually navigate. According to The Verge AI, Isabelle Reksopuro built an interactive map that tracks data center policy and public response across the world, pulling from Epoch AI’s dataset and legislation she scraped herself. The trigger was personal: Google’s sprawling campus in The Dalles, Oregon already drinks roughly a third of the city’s water supply, and the city quietly asked for 150 acres of Mount Hood National Forest to feed its growing demand.
What stands out here is how the project was built. The map uses Claude to search for new sources four times a day, check them against Reksopuro’s database, write fresh summaries, and update the feed automatically. “I wanted it to be self-updating, since I’m also a student,” she told The Verge AI. That’s a quiet but important data point about where AI tooling is going: a single undergrad can now run a global policy tracker that would have required a small newsroom five years ago.
The political map is weirder than you’d expect
The headline finding from Reksopuro’s work, as detailed in The Verge AI, is that opposition to data centers is one of the rare issues uniting Americans across party lines. After construction wraps, the facilities leave behind few permanent jobs and push regional power costs to record highs, per Bloomberg reporting cited in the piece. But the policy response is all over the place:
- Maine passed the first state-level moratorium on hyperscale data centers in April. Governor Janet Mills vetoed it.
- Texas hands out more than $1 billion a year in tax breaks to attract them, according to The Texas Tribune.
- Oregon is stuck in a fight over whether a city is really asking for water for its 16,010 residents or for a hyperscaler that won’t be named on the paperwork.
That patchwork is the real story. Hyperscalers can shop jurisdictions the way retailers shop sales tax rates, and most communities don’t know a facility is coming until the permits are already moving.
Why this matters now
The AI buildout is the largest infrastructure expansion of this decade, and the friction points are local: water, power, land, noise, property tax abatements. Reksopuro’s framing is sharper than most policy commentary on the topic. She isn’t anti-data-center. Her argument, in The Verge AI’s account, is that opacity is the problem.
“Right now, it’s this really opaque thing, and all of a sudden, there’s a facility. I think that if people knew about data centers beforehand, it would give them leverage. They would be able to negotiate: ask for job training programs, tax revenue, environmental monitoring, things to improve their community.”
That’s a useful reframe for anyone working in or around AI infrastructure. The question isn’t whether the buildout happens. It’s whether the deals get negotiated in public or in a closed session of a city council most residents have never attended.
Practical takeaways
For AI practitioners and the companies hosting compute:
- Expect site selection to get political fast. Maine’s moratorium attempt won’t be the last. Even vetoed bills set the terms of the next election cycle.
- Transparency is becoming a competitive moat. Operators who publish water, power, and tax-abatement data up front will face less organized resistance than those who route through shell entities.
- Local journalism and student-built trackers are now part of your stakeholder map. A map like Reksopuro’s can move from curiosity to a county commissioner’s reading list in a single news cycle.
For businesses planning AI workloads, the second-order effect is power pricing. If more states follow Maine’s lead, capacity in cheap-power markets gets scarcer and more expensive. The arbitrage window on “build anywhere” is closing.
More details on the map and the policy fights it tracks are at the original source.