The News
Constellation Energy Group CEO Joseph Dominguez said Tuesday the US is “very behind” China in the race to build up energy to feed AI data centers.
If building more power plants as quickly as China is the strategy to keep pace with the AI infrastructure expansion, “we’re in some trouble,” Dominguez said at Semafor World Economy in Washington, DC on Tuesday.
“If this is going to be a race between China and the US to build energy, might as well call it a day,” he said, noting that since 2010, China has built the entirety of the US electric system plus 50% more in terms of capacity.
But the US can catch up if it uses its grid more efficiently, Dominguez said. The US is now less industrialized, which means power demand is no longer around the clock, compared to China, he noted.
“Their system is running pretty much 24/7 at high capacity,” Dominguez said. “Most of the time in the country, most of the power plants, most of the lines, most of the substations aren’t being used … So if we could manage the peak energy demands, we [would] have a lot of slack in the system we could take advantage of.”
It’s possible that American consumers could end up with lower energy prices if this strategy is successful, he said.
Asked about local opposition to data center construction, Dominguez said the US will lose the AI race and risk national security if it doesn’t successfully compete with China.
“It’s imperative that we win this … for the defense of the nation and our way of life,” he said.
The US needs a national policy in place instead of state-by-state approvals for power projects, he said.
“If NIMBYism becomes the reason we lose the AI race, for whatever reason, we’re in a whole lot of trouble in this country,” he said.
Know More
Nuclear power has been growing more popular with Americans as an option for power buildup, Dominguez said. About 65% of Americans approve of nuclear power — the highest it’s been in 30 years, he said. That’s higher than offshore wind, while only 20% to 25% support data centers.
Constellation Energy Group is coming off an extremely busy time. It inked multibillion-dollar deals to supply electricity for AI data centers.
In January, it closed its $16 billion acquisition of Calpine Energy, which made it the nation’s largest competitive retail electricity supplier. It’s also the US’ largest producer of carbon-free electricity with a portfolio of wind, solar, hydro, and nuclear power.
Dominguez has previously argued that data centers can actually help bring the cost of electricity down if they curb their power intake during peak residential demand periods, as a way to address concerns about higher power bills due to data center usage.
“By making data center load more flexible and using targeted demand response for just a few critical hours each year, we can meet near‑term AI demand, maintain our competitive edge and keep energy costs stable — even as we continue building the next generation of power infrastructure,” Dominguez said in a LinkedIn post.