Trump sees an AI build-out that can deliver jobs, lower taxes and an edge over China; opponents see electricity costs, environmental strain and local decisions being swept aside. Even within his party, the political argument is shifting from whether to build to who pays the price.

At an August Cabinet meeting, Trump showcased the scale of the coming infrastructure by holding up an image of Meta’s Hyperion Data Center imposed over Manhattan. He has since cast data centers as a central economic and geopolitical bet, arguing that resistance hands an advantage to Beijing.¹

That message sharpened on Monday, when Trump told communities that rejecting the projects meant choosing to be “backwards and poor.” He urged them to “let Data Reign,” then warned: “If we kill the Golden Goose, you will only have yourselves to blame.”² For Trump, the calculation is blunt: accept the warehouses powering AI, gain jobs and investment; reject them, and China wins.

Residents confronting proposed sites have made a different calculation. They cite higher utility bills, intensive power and water use, environmental effects and the loss of local control. The backlash has become sufficiently fierce that officials in more than a dozen communities face recall efforts tied to approvals of the energy-hungry facilities.²

The split is now visible inside the administration and the GOP. Vice President JD Vance acknowledged that most opposition arises where a new data center means higher electricity prices, arguing that developers should add power to the grid rather than drain it: “If you build a data center, you should be putting power back into the grid, not taking it out.”³ Yet the White House’s cost-sharing pledge for tech companies remains voluntary.

Republican governors and candidates have also moved toward pauses, reviews and tighter incentives. In Ohio, the issue has become a campaign weapon: an NRSC memo warned that data centers are “the anchor hanging around Husted’s neck” in the Senate contest, a sign that Trump’s golden goose is becoming a midterm liability.⁴