AI is driving new energy "pragmatism," industry players say
Add Axios as your preferred source to
see more of our stories on Google.

Illustration: Annelise Capossela/Axios.
HOUSTON — The need to power the AI revolution is the top-of-mind issue for energy CEOs and thousands of professionals at the CERAWeek by S&P Global conference here.
Why it matters: Depending on who you listen to, AI will either lead to a decades-long extension in relatively unfettered fossil fuel use or may not cause electricity demand to skyrocket as much as many assume.
- In any case, it's leading to a new energy "pragmatism" or "realism," in which some are pushing climate change concerns to the back burner.
Zoom in: In presentations and hallway conversations, experts said that increasing electricity demand for AI-related data centers is partly driving a new "energy realism," as ADNOC CEO Sultan Ahmed Al Jaber said in a speech Tuesday morning.
- "The race for AI supremacy is essentially an energy play," Al Jaber stated.
- He said the AI race will be won by those with easy access to adequate supplies of affordable energy and supporting infrastructure.
- "We need every energy option available," Al Jaber said. "In fact, we need it all. We need an 'and, and' approach to meet the rapidly growing global energy needs."
The intrigue: Other energy leaders here have also stressed the already increasing energy demand from data centers, electrification and manufacturing.
- "We need more energy, we need more gas, we need more electricity," said TotalEnergies CEO Patrick Pouyanné .
- "Gas is going to be right at the forefront of meeting that power demand," he said.
Natural gas is a fossil fuel that tends to produce fewer greenhouse gases than coal but emits methane, a powerful global warming agent.
- Pouyanné said cutting natural gas' methane emissions is "fundamental" for continuing to rely on this energy source.
Reality check: Constellation Energy CEO Joseph Dominguez questioned the conventional wisdom about dramatically increased energy demand that may outstrip supply and require a new wave of natural gas and other fossil fuel energy sources.
- He noted that data center operators tend to shop the same data center to different parts of the country to determine where it could be built most cheaply and with particular energy sources.
- This can lead to a false assumption of many more new data centers for every one that actually gets built.
- "So I've got this a little bit upside down from other speakers you might have been hearing, you probably hear throughout the conference this kind of panic," Dominguez said. "I don't feel it."
Zoom out: Dominguez also cautioned against meeting the increased demand with carbon-intensive energy sources, emphasizing the very real climate change consequences of such a strategy.
- "We're going to have to address that issue, and we can't ignore it," he said of climate change.
- He presented an argument for investing in more carbon-free sources to power data centers, such as advanced nuclear technologies.
- Constellation already operates the largest fleet of nuclear power plants in the U.S.
Many data center users, such as tech giants like Microsoft, Google and Amazon, are searching for low-carbon energy options to power their data centers.
Between the lines: For some in the energy sector, like energy tech platform Kraken, there's no question that electricity demand will significantly increase in the U.S. and other countries for the first time in decades.
- This demand will be met with a mix of sources, including renewables, Kraken CEO Amir Orad told Axios.
- In addition, AI can help improve the energy efficiency of data centers, helping bring down the demand growth curve.
- "Technology is part of the solution, not the problem," Orad said.
What we're watching: The steps that major energy companies, utilities and other players take in prepping for the new AI era.
Go deeper: A deep dive into powering AI data centers
