Constellation Energy CEO Urges US to Follow China’s AI Power Strategy
US Must Follow China's Lead in Powering AI Data Centers, Says Constellation Energy CEO
As the world rapidly embraces artificial intelligence (AI), there’s one critical resource in short supply: power. According to Joe Dominguez, CEO of Constellation Energy Corp., the United States must learn from China to meet the growing electricity demands generated by AI technology. His comments come on the heels of a landmark deal with Microsoft Corp. to reopen the long-closed Three Mile Island nuclear plant, underscoring the role of nuclear power in sustaining the future of AI.
The Growing Power Demand of AI
Artificial intelligence is a game-changer across various industries, but it’s also a major energy consumer. AI data centers, which are critical to supporting cloud computing, machine learning, and other advanced applications, require enormous amounts of electricity. The power demand for these facilities is pushing the limits of traditional energy sources, forcing companies to rethink how and where they get their power.
Big tech companies are now planning data centers so large that they could only operate if they’re built right next to power plants. “We are seeing customers looking at multi-gigawatt data centers,” Dominguez stated in an interview. “These centers can only be powered if located where the electricity is produced.”
Why Renewable Energy Isn’t Enough
While renewable energy—such as wind and solar—has made strides, it cannot reliably support the power demands of massive AI data centers on its own. Solar and wind are intermittent, meaning they don’t always generate power when it’s needed most. To fill the gaps, energy companies are extending the lifespan of coal plants and ramping up plans for new natural gas generation. However, for many tech companies, nuclear energy offers the best solution, providing consistent and reliable power 24/7.
Adopting China’s Energy Model for AI
Dominguez points to China as a blueprint for how the US can handle the energy demands of AI. China is already taking steps to build AI data centers directly next to power plants, ensuring an uninterrupted energy supply. This approach contrasts sharply with the US, where long-distance transmission lines are often used to transport electricity to data centers. However, a shortage of these transmission lines, combined with the time it takes to connect new facilities to the grid, is creating delays. For data center operators, those delays are costly, especially given AI's growing importance to national security and the economy.
Constellation’s Role in Powering AI
Constellation Energy is positioning itself at the forefront of AI’s energy needs. The company recently announced a major agreement with Microsoft to restart a reactor at the Three Mile Island nuclear plant in Pennsylvania. The reactor’s power will be directed to Microsoft’s AI data centers, reinforcing the notion that nuclear power is a critical element in supporting AI operations.
This isn’t an isolated event. Earlier this year, Amazon announced a $650 million investment to acquire a data center campus connected to another nuclear plant in Pennsylvania. These developments signal that tech giants are increasingly looking to nuclear energy to power their massive AI data centers.
Energy Proximity: A Strategic Necessity
Locating data centers near power plants is more than just convenient—it’s becoming a necessity. AI-driven data centers consume vast amounts of electricity, often requiring multi-gigawatt levels of power. The closer a facility is to its power source, the more reliable and immediate the energy supply becomes. This strategy helps companies avoid the costly delays associated with building new transmission lines or waiting for grid connections.
The National Security Imperative
AI is not just an economic asset; it’s a key factor in national security. Dominguez highlighted this point, stating that the US must be successful in AI to remain competitive globally, both geopolitically and economically. Reliable power is crucial to that success, and delays in grid connection could hinder AI development at a critical time.
The Revival of Nuclear Energy
Nuclear power is experiencing a resurgence as AI continues to drive up energy demand. Nuclear reactors, which run continuously and produce vast amounts of electricity, are now viewed as a pivotal solution for data centers. For years, nuclear energy had been sidelined in favor of renewable sources, but its reliability makes it the perfect fit for the always-on demands of AI infrastructure.
Nuclear’s role isn’t just about data centers. Across the US, energy companies are increasingly relying on nuclear power to meet growing demands from electric vehicles, manufacturing facilities, and other power-hungry industries. The reopening of the Three Mile Island plant is a testament to nuclear’s comeback as a cornerstone of modern energy strategy.
Learning from China’s AI Power Strategy
China has made great strides in AI, not just by advancing its technology but by building the infrastructure needed to power it. China’s strategy of building data centers directly next to power plants gives the country a competitive edge in AI development. This proximity ensures immediate access to electricity without the delays caused by grid connection issues.
Dominguez argues that the US needs to follow a similar approach. Building AI data centers near power plants would provide the immediate, reliable energy that AI requires. It would also eliminate the need for long-distance transmission lines, which are in short supply and often take years to complete.
The Future of AI and Energy in the US
As AI continues to grow in importance, energy companies like Constellation are leading the charge to adapt to its demands. Nuclear power, long considered a relic of the past, is now a critical player in the energy mix. With companies like Microsoft and Amazon investing heavily in nuclear-powered data centers, it’s clear that nuclear energy will play a major role in AI’s future.
The question now is whether the US can scale its energy infrastructure quickly enough to keep up with the demands of AI. Following China’s model of building data centers near power plants may be the key to ensuring the country remains a global leader in AI development. The stakes are high—not just for tech companies, but for the country’s national security and economic future.
Conclusion
As AI transforms the global landscape, its growing demand for electricity is pushing the limits of existing energy infrastructures. Joe Dominguez’s call for the US to adopt China’s model of building data centers near power plants underscores the urgency of addressing these challenges. With nuclear energy making a comeback and companies like Microsoft and Amazon leading the charge, the future of AI power is taking shape. But to keep pace with global competitors like China, the US must rethink its energy strategies—and fast.