The rapid expansion of artificial intelligence infrastructure in the United States is facing a growing electricity challenge, with data centre power requirements expected to rise sharply over the next few years.
Morgan Stanley Research estimates that US data centres will require 97 gigawatts of new power between 2026 and 2028. Against that requirement, about 21 gigawatts are associated with data centres already under construction, while another 19 gigawatts of grid capacity is available, leaving an initial potential gap of 57 gigawatts.
The pressure is reflected in projections for data centre IT power demand. Installed IT power demand is expected to rise from 9.19 gigawatts in 2025 to 17.96 gigawatts in 2026, representing an increase of about 95 per cent. The figure is projected to reach 35.46 gigawatts in 2027, another increase of roughly 97 per cent, before reaching 52.31 gigawatts in 2028 and 78.57 gigawatts by 2029.
That would put 2029 demand at about 8.5 times the 2025 level, highlighting how quickly electricity requirements are increasing as AI infrastructure expands.
The changing design of AI servers is one factor behind the rise. More powerful systems based on Nvidia’s Vera Rubin and Rubin Ultra architectures are expected to push data centre IT power demand substantially higher as newer generations of computing equipment are deployed.
The issue is not simply the amount of electricity required, but how quickly new power can be brought online. Data centre projects can take years to connect to the grid, while AI companies and cloud providers are expanding computing capacity at a much faster pace.
Morgan Stanley’s latest assessment puts the initial 57-gigawatt gap before additional alternatives are considered. The research also points to measures including natural gas turbines, fuel cells and nuclear power infrastructure as possible ways of reducing the remaining deficit. Even with such measures, the estimated median net shortfall is about 33 gigawatts, according to reporting on the research.
The growing demand is already influencing discussions around where new data centres can be built and how quickly they can receive electricity. Access to available generation and grid connections is becoming an increasingly important consideration alongside land, cooling systems, networking and computing hardware.
For the AI industry, the figures highlight a practical constraint on further expansion. Building more powerful models requires increasingly capable servers, while those systems require more electricity to operate.
The challenge also extends beyond individual technology companies. Utilities, grid operators, governments and data centre developers will need to coordinate around generation, transmission and connection capacity if projected demand is to be met.
The scale of the expected increase means that electricity supply is becoming a central part of the AI infrastructure discussion. The pace at which new power capacity can be secured could influence how quickly planned data centre projects move from announcements to operation.
Dear Reader,
Ledger Life is an independent platform dedicated to covering the Internet Computer (ICP) ecosystem and beyond. We focus on real stories, builder updates, project launches, and the quiet innovations that often get missed.
We’re not backed by sponsors. We rely on readers like you.
If you find value in what we publish—whether it’s deep dives into dApps, explainers on decentralised tech, or just keeping track of what’s moving in Web3—please consider making a donation. It helps us cover costs, stay consistent, and remain truly independent.
Your support goes a long way.
🧠 ICP Principal: ins6i-d53ug-zxmgh-qvum3-r3pvl-ufcvu-bdyon-ovzdy-d26k3-lgq2v-3qe
🧾 ICP Address: f8deb966878f8b83204b251d5d799e0345ea72b8e62e8cf9da8d8830e1b3b05f
Every contribution helps keep the lights on, the stories flowing, and the crypto clutter out.
Thank you for reading, sharing, and being part of this experiment in decentralised media.
—Team Ledger Life
Community Discussion