A new study on the rapid construction of the data centers suggests that the sudden demand for electricity could drive up energy costs for everyone and harm the environment. This study was published on Monday by the Union of Concerned Scientists, and it says building data centers faster than utilities can supply electricity will lead to higher costs for everyone.
“They’re increasing the demand faster than you can increase the supply. How’re you going to do that?” said Mike Jacobs, a senior manager of energy for the organization.
The report, “Data Center Power Play,” models several electricity demand scenarios over the next quarter century, along with different energy policy approaches for meeting those demands. The aim of this is to estimate trillions of dollars in potential electricity, climate and health costs.
“With clean energy policies, we can do this cheaper with less air pollution, less health impacts, less health costs, than doing this business as usual,” Jacobs said. He added that building out an electric grid to fulfill the demand for power will take much longer than building a data center.
“This is a collision between the people whose philosophy is ‘move fast and break things,’ with the utility industry that has nobody that says move fast and break things,” Jacobs said, in reference to data centers owners’ rapid construction of the facilities. He added that future data center demand is hard to predict, thanks in part to limited disclosure from utilities and major tech companies, according to the report. How that demand is met will be critical for both public health and the climate.
READ: OpenAI to cover rising power costs for touted Stargate data centers (January 22, 2026)
“This is really a great moment for regulators to do what’s within their authority and sort out and assign the costs to those who cause them, which is an essential principle of utility ratemaking,” he said.
Tech companies have been on a data center-building spree of late, hoping to profit from the AI boom. Companies like OpenAI, Google, Meta, Amazon have been investing massively in data centers and projects like Stargate have been the backbone for AI infrastructure.
While data centers bring in jobs and digital upgrades, there have been serious concerns about the substantial energy and water they demand.
Aside from energy, the demand for water has been a matter of major concern. Data centers rely heavily on water-intensive cooling systems. A single 100 MW data center can consume over two million liters of water per day, roughly equivalent to the daily use of 6,500 households. In regions already facing water scarcity—like parts of Georgia, Texas, Arizona, and Oregon—this demand compounds stress on aquifers and municipal supplies.