AI companies are buying up land for billions, but the real world is starting to push back

Venture capitalists are investing billions in the next wave of artificial intelligence, data centers are being built worldwide — but the real world is increasingly answering with a resounding no. A story from Kentucky has become a symbol of this confrontation. One of the leading AI companies offered $26 million to an 82-year-old local resident for the right to build a data center on her land.

The woman refused. The company did not back down: now it is seeking rezoning of neighboring 2,000 acres to bypass one person's unwillingness. Private property, local communities, and the technological industry's appetite have collided head-on.

This case is not unique. As AI infrastructure moves beyond server racks and cloud providers into the physical world, resistance is growing. Data centers require vast tracts of land, water for cooling, and gigawatts of electricity.

Local authorities, farmers, and activists in an increasing number of U.S. states and European countries are asking uncomfortable questions: who pays for the strain on the power grid, who bears the environmental costs, and why are decisions being made without community input?

Against this backdrop, another notable event occurred in the industry: OpenAI quietly shut down Sora — its text-to-video generator. The product, which sparked a wave of excitement after its announcement in February 2024, never became mainstream. The closure of Sora raises a question that venture investors prefer not to voice aloud: the ability to effectively demonstrate technology and the ability to turn it into a sustainable business are completely different skills.

Venture funds, meanwhile, continue to place bets. Investment volumes in AI startups in 2024–2025 hit all-time records. Investors are convinced: the next wave — agent systems, physical AI, robotization — will deliver returns comparable to the internet boom of the early 2000s.

But the story of the Kentucky farmer reminds us: between promise and realization lies not only a technological but also a human barrier. AI infrastructure must physically exist somewhere — and that "somewhere" is always someone's land, someone's water, someone's sky. While the industry counts megawatts and funding rounds, the real world is beginning to count something else.