The Majors Move Forward to Fuel the Growing Demand of AI Data Centers

In 2023, the Scientific American published an article warning about the shocking amount of electricity that would be used by artificial intelligence (AI). The Scientific American referenced a peer-reviewed analysis published in Joule which predicted that trends in AI capacity were set to lead NVIDIA to shipping 1.5 million AI server units per year by 2027.¹ The 1.5 million servers would likely consume at least 85.4 terawatt-hours of electricity annually, which would be more than what many small countries use in one year.²

The Scientific American spoke with Alex De Vries, a data scientist at the Central Bank of the Netherlands and a Ph.D. candidate at Vrije University Amsterdam, about AI’s use of electricity. De Vries illustrated that “if you were to fully turn Google’s search engine into something like ChatGPT, and everyone used it that way — so you would have nine billion chatbot interactions instead of nine billion regular searches per day — then the energy use of Google would spike. Google would need as much power as Ireland just to run its search engine.”³ De Vries commented that although highly unlikely, “if we decide we’re going to do everything on AI, then every data center is going to experience effectively a 10-fold increase in energy consumption.”⁴

Alternatively, in an article published by Hannah Ritchie in Sustainability by Numbers, November 17, 2024, Ritchie conveyed that the projected growth of the electricity demand of AI data centers is lower than previously anticipated. Ritchie referenced a publication by the International Energy Agency (IEA) regarding the projected growth in the global electricity demand from 2023 to 2030.⁵ The IEA publication showed that data centers accounted for only 3% growth, as compared to the much higher rate of projected growth for electric vehicles and air conditioning, among others.⁶ Nevertheless, Ritchie agreed that energy demand is expected to grow just less than expected the previous year.⁷

In response to the growth in AI, two major oil and gas companies have revealed their plans to fuel AI data centers.⁸ In a news article published by Chevron on February 13, 2025, Chevron announced that it will tap into natural gas to help power AI data centers. Chevron announced that it will be working with GE Vernova (GE) and Engine No. 1, as they are aiming to deliver reliable energy to U.S. data centers.⁹ GE is expected to deliver seven gas-powered turbines to Chevron, which will first deliver power in 2027, helping provide energy that AI data centers need.¹⁰

Exxon is also planning to provide solutions for the increasing electricity demand derived from AI data centers. On December 11, 2024, Exxon published a news article outlining its plan.¹¹ Exxon plans to build a facility, detached from the existing grid infrastructure, that would use natural gas to generate a significant amount (>1.5 gigawatts) of high-reliability electricity for a data center.¹² Exxon plans to look for opportunities to source low-carbon intensity natural gas like the kind it is producing in the Permian Basin.¹³ Additionally, Exxon plans to use carbon capture to remove more than 90 percent of the associated CO2 emissions, then transport the captured CO2 to safe, permanent storage deep underground.¹⁴

Please stay tuned for future updates as the race to power AI data centers begins to unfold. Over the next few months, we will be learning more about how AI data centers will be powered and what sources of power provide the best option.

1. Lauren Leffer, The AI Boom Could Use a Shocking Amount of Electricity, Scientific American (October 13, 2023), https://www.scientificamerican.com/article/the-ai-boom-could-use-a-shocking-amount-of-electricity/.
2. Id.
3. Id.
4. Id.
5. Hanna Ritchie, What’s the impact of artificial intelligence on energy demand?, Sustainability by Numbers (November 18, 2024), https://www.sustainabilitybynumbers.com/p/ai-energy-demand.
6. Id.
7. Id.
8. US companies make bold move to power nation’s data centers, Chevron Corporation (February 13, 2025), https://www.chevron.com/newsroom/2025/q1/us-companies-make-bold-move-to-power-nations-data-centers.
9. Id.
10. Id.
11. Steel, ammonia and AI? Oh my! What can’t our CCS help decarbonize?, ExxonMobil Corporation (December 11, 2024), https://corporate.exxonmobil.com/what-we-do/delivering-industrial-solutions/carbon-capture-and-storage/steel-ammonia-ai-what-cant-ccs-help-decarbonize.
12. Id.
13. Id.
14. Id.