.jpg)
Copyright © 2026 by Enlit. All rights reserved. This article first appeared in January 2026 on Enlit.
However this collaboration to tackle the data centre boom must be done without punishing ratepayers, writes Hari Vasudevan.
JP Morgan recently announced that AI’s $5 trillion data centre boom will be felt in every debt market. In the energy sector, this comes as no surprise, with the forecast finances reflecting the already ballooning electricity demand brought by data centres.
In fact, global data center power demand is set to double in the next five years alone, driving nearly half of all new electricity demand growth this decade.
As it currently stands, the grid cannot keep up with this demand, which has prompted the hyperscalers behind the data centre boom to build their own power plants. But this doesn’t just carry consequences for utilities – ratepayers are forced to shoulder the costs.
In the US, utilities estimate they will invest at least $1 trillion in the next five years to power the country’s economic and energy security. And regulators are now confronting the question of who pays for it.
To keep pace and protect consumers in the process, utilities need to rewire their role, moving beyond simply providing energy to becoming strategic partners of the AI revolution.
The grid is showing its age: 70% of transmission lines and large transformers are more than 25 years old. Even when utilities are ready to invest, permitting, siting and interconnection reviews can drag out for years. Permitting bottlenecks are creating a hidden tax on ratepayers, and electricity rates are spiking at twice the rate of inflation.
Project-approval delays cost Americans tens of billions of dollars annually in higher energy bills, disproportionately hurting low-income families and minorities. Studies show that trimming federal permitting timelines by just one year could unlock $22 billion in additional returns on infrastructure investment. These permitting delays drive up wholesale prices, which inevitably land on the shoulders of ratepayers.
Some hyperscalers are turning away from the grid, pursuing a ‘bring your own power’ approach in the form of gas turbines, fuel cells, or even SMRs. — Hari Vasudevan
The clock is ticking, and hyperscalers aren’t going to hold out while utilities try to play catch-up. Some hyperscalers are turning away from the grid, pursuing a ‘bring your own power’ approach in the form of gas turbines, fuel cells, or even small modular nuclear reactors to independently fulfill their electricity needs.
But that comes with tradeoffs, including higher cost per kilowatt, greater carbon footprints, lower efficiency, unproven reliability, and reduced resiliency due to being isolated from the grid.
It’s not sustainable in any shape or form for the future of grid growth to solely rest on utilities’ balance sheets. Now, with hyperscale data centre demand booming, utilities have the opportunity to evolve from the primary financiers of new capacity to strategic partners. Deeper integration for the hyperscalers, not isolation, is the better option all around.
An alternative co-development approach is about passing the capital risk to the hyperscalers behind the demand. It gives utilities much-needed breathing room to focus on providing both data centers and households with sufficient grid interconnection and oversight.
In this model, utilities work with data centre hyperscalers to design and connect on-site or near-site generation, whether that’s natural gas turbines, fuel cells, small modular nuclear reactors, solar microgrids, or battery storage.
Ideally, AI data centers, which are massive, 24/7 power consumers, would pay their full costs. Across the country, residential consumers already pay higher rates than commercial and industrial users.
If data centre expansion accelerates under discounted electricity rates, household bills will rise even faster, precisely when affordability is most strained for the average rate payer. Alternatively, if hyperscalers flexed or reduced demand during less than 1% of annual hours, it could meaningfully reduce cost burdens for other utility ratepayers.
With data centers on the scene, the AI era ultimately demands collaboration that balances innovation with prudence in a unified front between hyperscalers and utilities.
When utilities and hyperscalers coordinate through flexible interconnection and dynamic load-management agreements, the grid itself becomes the safety net for the hyperscalers.
Power can be rerouted where it’s needed most during peak demand, minimizing outages and stabilizing prices, despite the massive energy load that data centers require. A modern, connected grid, rather than fragmented private power islands, keeps electricity reliable, resilient, and affordable for data centres and households.
Ultimately, the investment risk sits with the hyperscaler. Utilities retain their role as grid stewards, planners, engineers, builders, and operators, without being forced to put extra burdens on ratepayers to foot the bill for new power generation. They’re not left in the dust by hyperscalers, either, but are keeping pace with them.
Crucially, risk transfer is at the crux of this way forward. Recent rumblings liken the AI boom to the days of the dot-com bubble. Utilities can protect ratepayers from potential over-building by requiring data centre developers to finance their own interconnection and dedicated power capacity.
At the same time, every additional large customer beyond data centers, such as manufacturing plants and hospitals, helps spread fixed network costs, primarily poles, towers, and wires, in an effort to reduce average rates for households and small businesses.
Meanwhile, federal permitting reforms could make this co-development partnership a faster and more affordable reality. The Department of Energy’s Coordinated Interagency Transmission Authorizations and Permits (CITAP) initiative now streamlines environmental reviews for transmission and interconnected projects through a two-year unified process.
When utilities pair these federal accelerators with private sector capital from hyperscalers, affordability improves for ratepayers. This co-development model also encourages better load forecasting and cost recovery discipline. Instead of speculatively expanding grid capacity, utilities can tie upgrades to binding, long-term interconnection contracts with hyperscalers.
These contracts would include cost-sharing and demand-management provisions so utilities and ratepayers wouldn’t bear the financial brunt of their presence. In turn, data centers gain faster access and regulatory certainty and utilities' energy know-how, while consumers benefit from cost stability and even potential reduction.
This is the foundation for a partnership that aligns incentives across hyperscalers and households. Hyperscalers get reliability and speed, utilities capital relief, infrastructure modernization, and improved resiliency, regulators cost containment and transparency, and American households and small businesses are spared from runaway rates.
With data centers on the scene, the AI era ultimately demands collaboration that balances innovation with prudence in a unified front between hyperscalers and utilities.
The utility providers that master this balancing act will redefine themselves not as commodity suppliers, but risk-smart orchestrators of the emerging AI-energy ecosystem. They’ll be able to keep the lights on for everyone without having ratepayers foot the bill for the next infrastructure bubble.
Link to the original article: here
Copyright © 2026 by Enlit. All rights reserved. This article first appeared in January 2026 on Enlit.
However this collaboration to tackle the data centre boom must be done without punishing ratepayers, writes Hari Vasudevan.
JP Morgan recently announced that AI’s $5 trillion data centre boom will be felt in every debt market. In the energy sector, this comes as no surprise, with the forecast finances reflecting the already ballooning electricity demand brought by data centres.
In fact, global data center power demand is set to double in the next five years alone, driving nearly half of all new electricity demand growth this decade.
As it currently stands, the grid cannot keep up with this demand, which has prompted the hyperscalers behind the data centre boom to build their own power plants. But this doesn’t just carry consequences for utilities – ratepayers are forced to shoulder the costs.
In the US, utilities estimate they will invest at least $1 trillion in the next five years to power the country’s economic and energy security. And regulators are now confronting the question of who pays for it.
To keep pace and protect consumers in the process, utilities need to rewire their role, moving beyond simply providing energy to becoming strategic partners of the AI revolution.
The grid is showing its age: 70% of transmission lines and large transformers are more than 25 years old. Even when utilities are ready to invest, permitting, siting and interconnection reviews can drag out for years. Permitting bottlenecks are creating a hidden tax on ratepayers, and electricity rates are spiking at twice the rate of inflation.
Project-approval delays cost Americans tens of billions of dollars annually in higher energy bills, disproportionately hurting low-income families and minorities. Studies show that trimming federal permitting timelines by just one year could unlock $22 billion in additional returns on infrastructure investment. These permitting delays drive up wholesale prices, which inevitably land on the shoulders of ratepayers.
Some hyperscalers are turning away from the grid, pursuing a ‘bring your own power’ approach in the form of gas turbines, fuel cells, or even SMRs. — Hari Vasudevan
The clock is ticking, and hyperscalers aren’t going to hold out while utilities try to play catch-up. Some hyperscalers are turning away from the grid, pursuing a ‘bring your own power’ approach in the form of gas turbines, fuel cells, or even small modular nuclear reactors to independently fulfill their electricity needs.
But that comes with tradeoffs, including higher cost per kilowatt, greater carbon footprints, lower efficiency, unproven reliability, and reduced resiliency due to being isolated from the grid.
It’s not sustainable in any shape or form for the future of grid growth to solely rest on utilities’ balance sheets. Now, with hyperscale data centre demand booming, utilities have the opportunity to evolve from the primary financiers of new capacity to strategic partners. Deeper integration for the hyperscalers, not isolation, is the better option all around.
An alternative co-development approach is about passing the capital risk to the hyperscalers behind the demand. It gives utilities much-needed breathing room to focus on providing both data centers and households with sufficient grid interconnection and oversight.
In this model, utilities work with data centre hyperscalers to design and connect on-site or near-site generation, whether that’s natural gas turbines, fuel cells, small modular nuclear reactors, solar microgrids, or battery storage.
Ideally, AI data centers, which are massive, 24/7 power consumers, would pay their full costs. Across the country, residential consumers already pay higher rates than commercial and industrial users.
If data centre expansion accelerates under discounted electricity rates, household bills will rise even faster, precisely when affordability is most strained for the average rate payer. Alternatively, if hyperscalers flexed or reduced demand during less than 1% of annual hours, it could meaningfully reduce cost burdens for other utility ratepayers.
With data centers on the scene, the AI era ultimately demands collaboration that balances innovation with prudence in a unified front between hyperscalers and utilities.
When utilities and hyperscalers coordinate through flexible interconnection and dynamic load-management agreements, the grid itself becomes the safety net for the hyperscalers.
Power can be rerouted where it’s needed most during peak demand, minimizing outages and stabilizing prices, despite the massive energy load that data centers require. A modern, connected grid, rather than fragmented private power islands, keeps electricity reliable, resilient, and affordable for data centres and households.
Ultimately, the investment risk sits with the hyperscaler. Utilities retain their role as grid stewards, planners, engineers, builders, and operators, without being forced to put extra burdens on ratepayers to foot the bill for new power generation. They’re not left in the dust by hyperscalers, either, but are keeping pace with them.
Crucially, risk transfer is at the crux of this way forward. Recent rumblings liken the AI boom to the days of the dot-com bubble. Utilities can protect ratepayers from potential over-building by requiring data centre developers to finance their own interconnection and dedicated power capacity.
At the same time, every additional large customer beyond data centers, such as manufacturing plants and hospitals, helps spread fixed network costs, primarily poles, towers, and wires, in an effort to reduce average rates for households and small businesses.
Meanwhile, federal permitting reforms could make this co-development partnership a faster and more affordable reality. The Department of Energy’s Coordinated Interagency Transmission Authorizations and Permits (CITAP) initiative now streamlines environmental reviews for transmission and interconnected projects through a two-year unified process.
When utilities pair these federal accelerators with private sector capital from hyperscalers, affordability improves for ratepayers. This co-development model also encourages better load forecasting and cost recovery discipline. Instead of speculatively expanding grid capacity, utilities can tie upgrades to binding, long-term interconnection contracts with hyperscalers.
These contracts would include cost-sharing and demand-management provisions so utilities and ratepayers wouldn’t bear the financial brunt of their presence. In turn, data centers gain faster access and regulatory certainty and utilities' energy know-how, while consumers benefit from cost stability and even potential reduction.
This is the foundation for a partnership that aligns incentives across hyperscalers and households. Hyperscalers get reliability and speed, utilities capital relief, infrastructure modernization, and improved resiliency, regulators cost containment and transparency, and American households and small businesses are spared from runaway rates.
With data centers on the scene, the AI era ultimately demands collaboration that balances innovation with prudence in a unified front between hyperscalers and utilities.
The utility providers that master this balancing act will redefine themselves not as commodity suppliers, but risk-smart orchestrators of the emerging AI-energy ecosystem. They’ll be able to keep the lights on for everyone without having ratepayers foot the bill for the next infrastructure bubble.
Link to the original article: here
Hari Vasudevan, PE, is a serial entrepreneur and engineer focused on AI-driven solutions for utilities, construction, and storm response. As Founder and CEO of KYRO AI, he leads the development of AI-powered software that helps utility, vegetation, and field service teams digitize operations, improve storm response and restoration, and reduce operational risk. He also serves as Vice Chair and Strategic Adviser for the Edison Electric Institute’s Transmission Subject Area Committee and holds bachelor’s and master’s degrees in civil engineering with professional engineering licensure in multiple states.
.jpg)
.jpg)
.jpg)