Google flags decade-long wait to plug data centers into US grid as demand surges

• What happened: Google says connecting new US data centers to the electricity transmission system can now take 10+ years in some regions.
• Why it matters now: AI growth is accelerating power demand faster than the grid can handle.
• What changes for people: Slower cloud rollout, higher power costs and delays in digital services could follow.
• Who is affected: Big tech firms, utilities, regulators, and communities hosting energy-intensive data centers.

Google warns US grid delays are now its biggest hurdle

The race to expand artificial intelligence and cloud computing in the United States is running into a critical bottleneck. Google says long waits for transmission access have become the single biggest obstacle to powering new data centers, with some utilities citing interconnection studies lasting more than twelve years.
This matters because data centers are now core infrastructure for everyday software, online services and AI models that power both businesses and consumer products.

A growing gap between power demand and infrastructure

The issue reflects a widening mismatch between the speed of the tech sector and the pace of US energy planning. The US transmission grid moves electricity across states but is constrained by permitting delays, limited capacity and slow approval processes.

According to Marsden Hanna, Google’s Global Head of Sustainability and Climate Policy, the company has been told by at least one utility that simply studying a connection request could take over a decade. Hanna shared these remarks at an event hosted by the American Enterprise Institute, underscoring how long the queues have become.

Why the grid cannot keep up

Data centers require massive and continuous electricity loads. With AI training and cloud services accelerating, major tech companies are competing for scarce grid capacity.

Energy analysts say this has implications for:

• Economic growth as AI-driven industries scale

• National competitiveness in computing infrastructure

• Power prices and reliability, especially in high-demand states

• Climate goals, because cleaner energy sources must travel through congested transmission lines

Regulators from the Federal Energy Regulatory Commission (FERC) and regional grid operators have already acknowledged transmission queues as a structural risk for the US economy.

Google explores co-location to bypass delays

To sidestep the bottleneck, Google is exploring co-location, where a data center is built directly next to an existing power plant.
This can underlineallow companies to bypass the transmission network and long waitlists. However, Google said the goal remains to eventually connect such facilities to the grid, not operate them in isolation.

The model is controversial because it raises questions around:

• Who bears the cost of these arrangements

• How power is allocated away from existing customers

• Whether reliability is affected if a single user takes a large power share

Federal and regional regulators are now examining guidelines for co-location, signaling that new rules may be on the way as demand surges.

What experts say

Energy policy researchers note that the United States built transmission lines at a much faster pace in previous decades, but recent years have been slowed by permitting disputes, local opposition, and environmental reviews.

Utilities argue that grid upgrades require long-term planning, while technology firms argue the delays threaten innovation and investment.

Why this matters to readers

Even for consumers far from data centers, the impact is tangible. AI services, cloud tools, streaming platforms, and enterprise software all rely on data centers. Slower buildout could mean:

• Slower deployment of new digital services

• Higher energy costs passed through to businesses

• Regional investment shifts toward states willing to expand infrastructure.