The U.S. power grid has a timing problem. Electricity demand is accelerating quickly, driven in part by data center buildout. But the infrastructure needed to serve that demand moves slowly. New transmission lines can take a decade to permit and build. Distribution upgrades are lagging. Generation projects sit in interconnection queues for years. In some markets, the average time between a project requesting grid access and actually plugging in now stretches anywhere from three to eight years.

That mismatch is starting to shape the politics of data center growth across the country. Companies are eager to reap the benefits of AI, but Americans are increasingly wary of the costs: rising electricity bills, grid congestion, and infrastructure investments that appear to serve tech companies more than local residents.

So stakeholders are beginning to ask a simple question: If large loads need power quickly, shouldn’t they help the system that powers them?

Illinois offers a glimpse of how that conversation is evolving. Legislation under consideration would require large data centers to bring their own clean capacity equal to their maximum demand or to provide flexibility that reduces their impact on the grid. Importantly, the bill defines clean capacity broadly: it could include not just new wind, solar, and storage, but also demand response, energy efficiency, virtual power plants, and other demand-side technologies.

In other words, the policy explicitly recognizes that capacity doesn’t have to come from a new power plant. It can also come from reducing or shifting demand. And not just on-site demand, but by making investments in distributed energy resources that can support nearby communities.

That idea sits at the center of a recent report AHG put together with the Alliance to Save Energy. Instead of building new centralized generation to accommodate a data center, the large load funds investments in distributed energy resources and demand-side upgrades. By lowering peak demand in targeted locations on the grid through technologies like heat pumps, rooftop solar, batteries, and demand response, this approach creates the capacity headroom needed to serve the new facility.

The idea isn’t to power an entire hyperscale data center with residential solar panels. It’s to address the last slice of capacity that often determines whether a project can interconnect this year or wait several years for transmission upgrades. Think of it as the distributed version of “bring your own capacity.”

The strategy is already beginning to show up in real-world projects. Earlier this year, Google announced an agreement with Xcel Energy in Minnesota that will bring roughly 1.9 gigawatts of new clean energy onto the utility’s system. While many of these investments are utility-scale generation resources, Google is also funding investments in Xcel’s distributed battery network through its Capacity*Connect program, which aggregates smaller batteries across the system to increase available grid capacity. That approach reflects a broader realization emerging in the electricity sector: some of the fastest capacity the grid can access is already sitting inside homes and buildings.

Replacing inefficient electric resistance heating with modern heat pumps, for example, can dramatically reduce winter peak demand. Distributed batteries can shift electricity consumption away from peak hours. Smart thermostats and other flexible devices can reduce load exactly when the grid needs relief. Taken together, these resources start to look less like scattered efficiency programs and more like a new form of infrastructure.

Unlike most grid infrastructure, those investments deliver immediate benefits to households. A high-efficiency heat pump, especially when it replaces a low-efficiency solution like electric resistance heating, can cut a family’s electric bill by hundreds of dollars per year. Weatherization upgrades improve comfort while lowering energy consumption. Batteries and flexible demand technologies help stabilize the grid during extreme weather, offering an important resilience benefit for homeowners.

The community benefit piece matters more now than even a few years ago. Across the country, data center development faces growing local resistance. Residents worry about electricity prices, water use, and whether massive new facilities will strain infrastructure without delivering meaningful benefits in return. At least 25 U.S. data center projects were canceled in 2025 due to local community opposition, quadruple the number in 2024.

Distributed capacity offers a different model. Instead of arriving as a massive new electricity consumer, a data center can arrive alongside a wave of investments that make the local grid stronger and homes cheaper to power.

Utilities are increasingly interested in this approach as well. Many are searching for ways to manage rapid load growth without triggering massive capital spending on new infrastructure. Distributed resources can often be deployed far faster than new power plants or transmission lines, offering near-term capacity relief while long-term projects move through planning and permitting.

None of this happens without detailed planning, of course. Distributed capacity has to be deployed in the right locations to relieve actual grid constraints. Utilities need reliable ways to measure and verify the capacity value of demand-side resources. And programs need to be designed to serve as genuine community investments — not just tech companies buying goodwill.

But the direction of travel is becoming clearer. The traditional model of grid expansion — build generation, then transmission, then distribution — was designed for a world where electricity demand grew slowly and predictably.

The world we’re entering now looks very different. With demand rising quickly, and affordability concerns increasing, the most interesting question may not be whether data centers bring their own power. It's whether they bring energy solutions to everyone else along the way.

— Annie Gilleo, Vice President, AHG