Elon Musk has warned that the rapid expansion of artificial intelligence (AI) could soon hit a wall—not because of software or hardware limitations, but due to a looming shortage of electricity.
In an interview with CNBC on Tuesday, the billionaire founder of xAI said the industry may begin to feel the effects of a power generation crisis by mid to late 2025 as data centers demand more energy than the current U.S. grid can reliably supply.
Musk, who has previously sounded alarms over AI’s unchecked growth, said his AI startup xAI is now working to build a gigawatt-scale data center outside Memphis, Tennessee—a facility that will match the capacity of a standard U.S. nuclear power plant. He expects the center to be operational in six to nine months.
Register for Tekedia Mini-MBA edition 17 (June 9 – Sept 6, 2025) today for early bird discounts. Do annual for access to Blucera.com.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register to become a better CEO or Director with Tekedia CEO & Director Program.
“As we solve the transformer shortage, there will be the fundamental electricity generation shortage,” Musk said. “My guess is people are going to start hitting challenges with power generation maybe by the middle of next year, end of next year.”
The tech mogul identified three bottlenecks to scaling AI: high-performance chips, transformers (which modulate voltage from power plants to make it usable by computers), and ultimately, the availability of electricity itself.
Google Also Raised the Alarm Earlier
Musk’s concerns echo warnings raised by Alphabet’s Google in February. At a conference hosted by the Nuclear Energy Institute, Caroline Golin, Google’s global head of energy market development, said the company realized it could not sustain data center growth using existing grid capacity—especially as weather-dependent renewables like solar and wind create volatility.
“We ran into a very stark reality that we didn’t have enough capacity on the system to power our data centers in the short term and then potentially in the long term,” Golin said.
In response, Google began exploring nuclear energy options to ensure reliable, round-the-clock electricity for its expanding infrastructure. The tech giant is now among a growing number of companies considering partnerships with nuclear operators to safeguard their AI ambitions.
China Builds, U.S. Lags
According to Musk, while the United States grapples with stagnant electricity generation, China is charging ahead.
“China power generation looks like a rocket going to orbit and U.S. power generation is flat,” he told CNBC.
The contrast reflects broader geopolitical tensions around AI supremacy. As both countries race to lead in AI innovation, energy infrastructure is emerging as a critical frontier.
To bridge the gap in Memphis, xAI is deploying natural gas turbines to help power its Colossus data center. But the company’s approach has already stirred controversy. Environmental advocates accuse xAI of violating the Clean Air Act and operating as a “major source of air pollution” without necessary permits or mitigation technology.
The use of fossil fuel-based energy, particularly without adequate oversight, has raised concerns about the climate impact of powering AI. Critics say the AI gold rush could worsen emissions and undermine efforts to decarbonize the energy sector.
Utilities Divided on Demand Outlook
While utilities such as Dominion Energy report strong, sustained demand for data center power—especially in hotbeds like Northern Virginia—others urge caution.
Dominion, which serves the world’s largest data center market, told investors during its recent earnings call that demand is showing no signs of slowing down, despite broader macroeconomic jitters. The company has aggressively expanded its infrastructure to meet this need.
However, Constellation Energy, the largest operator of nuclear power plants in the U.S., isn’t as optimistic. CEO Joe Dominguez cautioned that some forecasts may be overly bullish as developers court multiple jurisdictions, leading to double-counting in demand estimates.
“I just have to tell you, folks, I think the load is being overstated,” Dominguez said during the company’s Q1 earnings call. “We need to pump the brakes here.”
However, Musk’s comments highlight a growing tension in the AI industry: the very technology that promises to reshape the world could be undermined by the infrastructure needed to sustain it.
Between soaring chip demand, constrained energy grids, and intensifying environmental scrutiny, companies at the forefront of AI face a balancing act. The challenge is no longer just about building smarter models—but about securing the physical power to run them. This situation has stirred questions on whether the U.S. grid can keep up with the AI age, or will the nation’s next tech revolution stall at the switch.