Open your electric bill lately and squint at the total? You're not imagining it — and in a growing number of places, one of the reasons is something you never plugged in: artificial intelligence.
The AI boom runs on data centers, and data centers run on staggering amounts of electricity. A recent analysis of the PJM grid — the market that serves all or part of 14 mid-Atlantic and Midwest states — found that surging power demand from data centers was a primary reason for roughly $23 billion in customer price increases, hikes expected to last through at least 2028. In other words, the cost of feeding AI is quietly landing on ordinary households.
Why AI is so hungry for power
Every AI chatbot answer, image generation, and model training run happens inside a data center packed with power-thirsty GPUs that run around the clock. Training a large model can consume as much electricity as hundreds of homes use in a year, and once trained, the model keeps drawing power every time someone sends it a prompt. Multiply that across millions of users and thousands of companies racing to deploy AI, and demand explodes.
Data centers already accounted for a meaningful and fast-growing slice of national electricity use before the AI wave — and AI has poured fuel on that trend. Utilities are now planning for demand growth they haven't seen in decades.
How that demand ends up on your bill
Here's the part that catches people off guard. You don't have to use AI to help pay for it. The mechanism is all about how grid costs get shared:
When a data center plugs into the grid, the utility often has to upgrade shared infrastructure — new substations, beefier transmission lines, and additional generating capacity. In principle, regulators try to charge costs to whoever causes them. But once an upgrade becomes part of the grid everyone relies on, its cost gets spread across all customer classes — including residential ratepayers who never asked for a data center next door.
On top of that, a surge in overall demand pushes up wholesale electricity and capacity prices for the whole region. So even setting aside the direct upgrade costs, the sheer scale of new AI demand tends to lift prices for everyone connected to the same grid.
The fight over who should pay
This is now one of the hottest questions in energy regulation. Consumer advocates argue that hyperscale data centers — backed by some of the richest companies on earth — should shoulder the full cost of the grid buildout they trigger, rather than socializing it onto households.
In response, several states and utilities are exploring special "large load" tariffs that make big data centers pay more of their own way: long-term contracts, minimum payment commitments, and rules that stick them with the cost of the specific infrastructure they require. Whether those measures move fast enough to protect ordinary ratepayers — and whether they can claw back the increases already baked in through 2028 — is very much an open question.
What it means for you
- Expect upward pressure on bills in regions with heavy data-center growth, especially across the PJM footprint, for the next few years.
- It's not evenly distributed. If a data-center cluster is landing near you, your local rates are more exposed.
- Watch your state regulator. Public utility commissions are where the "who pays" fight is being decided — and public comment periods actually matter.
- Efficiency still helps. You can't opt out of grid-wide increases, but trimming your own usage softens the blow.
The bottom line
AI feels weightless — a few words typed into a box. But behind that box sits a physical machine drawing real power from a real grid, and building out that grid costs real money. Right now, a chunk of that bill is quietly being shared with everyone who flips a light switch. The technology may be new, but the economics are old: when demand jumps and infrastructure has to catch up, somebody pays. Increasingly, that somebody is you.
Comments
Post a Comment