Mervin Raudabaugh rejected a roughly $15.7 million offer to sell his 261-acre Pennsylvania farm for an AI data center and accepted a $1.9 million conservation easement that keeps the land restricted to agricultural use. Similar refusals have occurred elsewhere, including a Kentucky family that turned down a $26 million bid for half of a 1,200-acre farm. Opponents cite concerns about noise, energy and water use, and air quality. Large AI data centers can occupy dozens to hundreds of acres and require substantial cooling and backup power infrastructure.
Pennsylvania Farmer Rejects Nearly $16M Offer for AI Data Center — Chooses Conservation
Jamesbrey/Getty Images
Mervin Raudabaugh declined an estimated $15.7 million offer to sell his 261-acre Pennsylvania farm for construction of an AI data center and instead accepted a voter-backed conservation easement that preserves the land for agriculture.
The developer's proposed purchase amounted to roughly $60,000 per acre. Raudabaugh opted for a conservation easement worth about $1.9 million — roughly $7,200 per acre — that leaves the property in private ownership but legally prohibits future commercial, industrial, or residential development.
Serg Grbanoff/Shutterstock
Why Landowners Are Pushing Back
Across the U.S., tech firms and data-center developers have been buying large rural parcels or seeking land via eminent domain to build expansive campuses that power artificial intelligence and cloud services. Neighbors and landowners often raise concerns about increased noise, higher utility demands, water consumption for cooling, and local air quality impacts tied to generators and backup power systems.
In one Kentucky example from April 2025, a family that owns a 1,200-acre farm turned down an anonymous $26 million offer for half their property — an offer that worked out to more than $43,000 per acre despite local average land values near $6,000 per acre. That family said it did not want to be the generation to convert a multigenerational farm into an industrial data center.
Gerville/Getty Images
What AI Data Centers Require
AI-focused data centers can require hundreds of acres to accommodate large buildings, extensive cooling and power infrastructure, and robust security. They house servers, networking equipment, and support systems such as HVAC, backup generators, and power distribution. At some sites — notably in parts of Northern Virginia — reports indicate use of many large generators, raising local concerns about noise and emissions.
Examples of large AI or hyperscale sites include xAI’s Colossus 2 in Shelby, Tennessee (reported to occupy about 100 acres with a building footprint near 1,000,000 square feet) and Riot Platforms’ Rockdale campus in Texas (roughly 200 acres, with about 430,000 square feet of computing space).
Landowners who choose conservation easements, like Raudabaugh, prioritize preserving farmland, protecting water resources, and maintaining community character over lucrative offers from data-center developers. Their decisions underscore the ongoing tension between rapid infrastructure expansion for AI and local efforts to safeguard rural landscapes and quality of life.
Help us improve.
