Big Tech executives from Amazon, Google, Meta, Microsoft, xAI, Oracle, and OpenAI are set to sign a White House pledge committing their companies to build independent power supplies for data centers, rather than drawing from public electricity grids. The move is intended to shield consumers from rising electricity costs driven by the rapid expansion of AI infrastructure. Experts say the goal is unlikely to be fully achievable.
President Donald Trump promoted the plan during his State of the Union address, telling Americans that “no one’s prices will go up” as a result of energy demand from AI data centers. Industry executives have already indicated the commitment will not be binding.
Why Consumer Bills Are Already Rising
Residential electricity costs rose 6 percent nationwide in February compared with a year earlier, according to the US Energy Information Administration. States with high concentrations of data centers saw sharper increases: New Jersey reported a 16 percent rise, Pennsylvania a 19 percent rise. Natural gas prices, extreme weather, and decades of underinvestment in grid infrastructure have all contributed, making it difficult to isolate any single cause.
US data center power demand is projected to more than triple by 2035, climbing from nearly 35 gigawatts in 2024 to 106 GW, according to BloombergNEF. That scale of growth puts pressure on the grid regardless of how new facilities connect to it.
“Regardless of how these data centers connect, behind the meter or as part of the network, you’re going to increase demand,” said Ari Peskoe, director at Harvard Law School’s Electricity Law Initiative.
The Supply Chain Problem
Nearly three-quarters of planned generation equipment for data centers runs on natural gas, according to energy research firm Cleanview, which is tracking 56 GW of projects across the US. The problem is that gas turbines are already in critically short supply, with wait times stretching as long as seven years for new orders.
Turbine manufacturer GE Vernova announced plans to expand production by 25 percent. Mitsubishi Power said it would double output over the next two years. Both moves may fall short of demand. Two-thirds of gas projects currently in development across the US have not yet announced a turbine manufacturer, according to Global Energy Monitor.
Rising turbine prices create a secondary effect. As tech companies compete aggressively for the same equipment, utilities and industrial customers face higher costs for their own generating capacity, which can ultimately be passed on to ratepayers. The pledge’s consumer-protection logic has a built-in leak.
Interim Fixes and Their Limits
To bridge supply gaps, data center operators are turning to reciprocal engines and diesel generators. Both options, along with standard gas turbines, are not engineered for the continuous, uninterrupted power loads that data centers require. Google and Microsoft have also struck deals to restart nuclear power plants, but those projects are measured in years, not months.
Josh Price, director of energy and utilities at strategy firm Capstone, described the pledge candidly: Big Tech is “trying to push back against the narrative that they’re the bad guy.”
The political intent behind Wednesday’s signing is clear. The practical barriers to delivering on it are equally clear. Supply chains for power generation equipment were already strained before the AI buildout accelerated. The commitment tech companies are making at the White House may define the ambition without resolving the underlying math.
Photo by Leif Christoph Gottwald on Unsplash
This article is a curated summary based on third-party sources. Source: Read the original article
