AI demand – PG&E entered 2026 expecting a year of new electricity demand—then nearly all of it was spoken for within two months. With AI data centers, EVs, and new factories all pulling power from an aging network, utilities face interconnection queues that can stretch ye
For PG&E, 2026 didn’t start with a gradual climb. It started with a backlog—one that arrived faster than the system designed to absorb it.
The utility entered 2026 expecting a year’s worth of new electricity demand. Barely two months later, nearly all of it was already spoken for. Interconnection requests piled up faster than planners had expected, overwhelming a regulatory system built for a time when electricity demand barely moved.
That world is gone.
Load growth that historically ran below 1% annually hit 4% at some grid operators last year. according to a report by the Lawrence Berkeley National Laboratory. Bain and Company projects that AI data centers alone could consume up to 9% of total U.S. electricity by 2030—adding more than 150 terawatt-hours of demand the current grid was never really built to handle. A third of that new demand is concentrated in Virginia. Texas. and California. according to Pew Research Center. placing extraordinary pressure on regional systems already straining to keep up.
AI is what people notice first, but it isn’t the only new load. EVs. new factories. and industries shifting off diesel and gas are all drawing from the same aging grid at the same time. At Southwest Power Pool. which oversees electricity across 17 states. officials compared last year’s surge in demand to two large nuclear plants suddenly appearing on a grid with roughly 56 gigawatts of capacity.
By late 2024, more than 2,600 gigawatts of proposed generation and storage projects were waiting to connect to the grid, according to Lawrence Berkeley National Laboratory—more than twice the country’s entire installed capacity.
PG&E’s pressure point is bluntly shared across the industry. David Sawaya. PG&E’s director of rate-reducing load growth. says virtually every utility he knows has seen its interconnection queue swell by 50% to 150% in just two years. “The process does not move at the speed of business,” he tells Fast Company. “And right now, the business is moving very fast.”.
A decade behind, trying to catch up
Utilities aren’t exactly starting from strength. Carlos Elena-Lenz, who leads digital enablement at Hitachi Energy, recalls a colleague joining the company after decades in oil and gas offering a blunt assessment: utilities globally are about a decade behind that industry in adopting AI.
The culture inside many utilities. Elena-Lenz says. has been built around what he calls a “break-fix” model—waiting for equipment to fail rather than predicting and preventing failure. Many utilities, he adds, still cannot tell customers with GPS-level precision where their own assets sit on the grid.
There’s also the data problem. Yuriy Yuzifovich. CTO of AI at GlobalLogic. a Hitachi subsidiary that builds digital systems for energy companies. says utilities often want AI systems their infrastructure cannot support. Equipment across the grid is generating useful data. but much of it never reaches the enterprise systems designed to use it. Even when data does make it through, utilities often lack the people or workflows needed to act on it.
“Intelligence is just the tip of the iceberg,” he says. The larger challenge lies beneath it: rebuilding data infrastructure, changing internal processes, and retraining workforces to actually use these systems effectively.
Each new AI workload arriving on the grid creates pressure across three dimensions at once: more power, more cooling, and more filtration. For an industry already a decade behind, that’s catching up on multiple fronts—at the same time.
Still, some progress is visible. Utilities that spent years asking what they should avoid are now asking what they can do. GlobalLogic is deploying AI systems inside utility operations that continuously predict where grid stress will appear hours before it materializes. Hitachi Digital has begun using synthetic data to replicate entire power-grid networks for testing. allowing models to be stress-tested against simulated conditions before touching live infrastructure.
The regulatory wall
Tools to modernize the grid are accelerating, but the rules governing them often lag behind.
In conversation. Sawaya describes the California regulatory process as cumbersome rather than broken: a deliberate. stakeholder-driven structure built for a different era. Bringing a new proposal to the California Public Utilities Commission. gathering testimony. building a record. and securing a decision can take two years.
For a data center developer working on a business timeline, that can kill a project before it properly begins.
Richard Schomberg. special envoy for smart electrification at the International Electrotechnical Commission (IEC). puts a harder number on the broader bottleneck. Interconnection timelines across the U.S. can stretch to seven years, constrained by transmission capacity, substation readiness, and a queue system designed for far lower volumes.
Jesse Jenkins. who leads the ZERO Lab at Princeton University. argues the industry may be thinking about the problem in the wrong terms. “It doesn’t make sense to build a whole new transmission line that sits there 365 days a year when you only need it for a few hours a year. ” he told Bloomberg News late last year. His research suggests that data centers willing to reduce their load during peak demand hours could connect to the grid years earlier—and at significantly lower cost—than those demanding guaranteed full power around the clock.
PG&E has responded with reforms worth watching. A proposal called Rule 30. approved by the California Public Utilities Commission in July 2025. requires large customers connecting to the transmission system to pay their full interconnection costs upfront. with reimbursement tied to the revenue their electricity consumption generates over time. If a data center commits. pays. and later scales back its power usage. minimum-demand fees ensure the shortfall does not fall on existing customers.
At Southwest Power Pool. a consolidated planning process pending federal approval promises to cut interconnection study timelines from more than a year to six months. A separate fast-track program. recently approved by federal regulators. can bring certain large-load and generation projects online in as little as 90 days.
For all the momentum, the fixes are still uneven—one utility in one state is still a long way from a national policy response.
Building around the grid
Some companies are moving faster than the grid they need.
Charlotte Meerstadt. founder and CEO of Fram Energy. is building the financial infrastructure for an energy economy that is reorganizing itself. The shift from centralized to decentralized energy generation—where businesses generate their own power and sell it directly to buyers through long-term agreements—is growing at a compound annual rate of 30%. Meerstadt says.
Her explanation is straightforward: rising energy costs, unreliable grid supply, and the appeal of locking in a fixed electricity rate for 20 to 25 years instead of absorbing whatever the market delivers.
But the billing infrastructure for these deals barely exists. Independent power producers. including solar farms. storage operators. and property owners selling power through direct agreements. typically manage billing through spreadsheets—manually shuttling data between operations and accounting teams.
Meerstadt describes what that looks like in practice: one of Fram Energy’s customers had been underbilling by $150,000 over six months without realizing it, and another was spending $50,000 a month simply to get invoices out correctly and on time.
Fram’s platform automates the process, handling more than 200 variables per bill across thousands of monthly transactions. It also shows buyers exactly what they would have paid at standard grid rates versus what they actually paid—an accounting comparison most independent producers cannot perform on their own.
Meerstadt calls it the “Stripe for the decentralized energy future.” She projects the decentralized energy market could reach $2 trillion by 2034.
Who pays for the upgrade
As the race for electricity intensifies, the debate is turning to a familiar question—only sharper: who pays.
Schomberg is direct about where the cost burden sits today. Tech companies are capturing enormous economic value from AI infrastructure while the capital costs required to enable that infrastructure are being socialized across millions of ratepayers who receive none of the upside.
He advocates for a “cost causation” principle: those whose demands drive the investment should bear its cost. PG&E’s Rule 30 moves in that direction. but applying the same logic nationally would require coordinated policy response the U.S. energy regulatory system has not historically been built to deliver.
Rob Gramlich, president of the independent power sector consultancy Grid Strategies, warns that politics doesn’t wait for engineering. “I don’t think we’ve seen the end of the political repercussions,” he told CNBC in January. “And with a lot more elections in 2026 than 2025, we’ll see a lot of implications.”.
Retail electricity prices have already been rising faster than inflation since 2022 and are forecast to climb another 5% this year, according to a short-term energy outlook by the U.S. Energy Information Administration.
For families and small business owners already stretched thin, those numbers aren’t abstract. They’re the moment where a policy choice turns into a household bill.
Hitachi, GlobalLogic, PG&E, Southwest Power Pool, and Fram Energy are each tackling a piece of a large, fast-moving problem. But the grid is still a shared resource, and the question of who bears the cost of transforming it is as political as it is technical.
For now, the debate has only just started.
PG&E AI data centers power grid interconnection queues Lawrence Berkeley National Laboratory Bain and Company Pew Research Center Southwest Power Pool California Public Utilities Commission Rule 30 Hitachi Energy GlobalLogic synthetic data International Electrotechnical Commission Princeton ZERO Lab Fram Energy decentralized energy Grid Strategies U.S. Energy Information Administration retail electricity prices