Chevron chief executive Mike Wirth says artificial intelligence could reshape the energy industry for decades, as power-hungry data centres force technology companies to secure reliable electricity supplies on an unprecedented scale.
BRUSSELS, August 6, 2026 — The artificial-intelligence investment boom is rapidly becoming an energy story—and Chevron CEO Mike Wirth believes the industry must build electricity generation and infrastructure quickly enough to prevent power shortages from slowing the expansion of data centres.
Speaking in an interview with Fox Business, Wirth discussed Chevron’s twenty-year agreement to supply electricity to a Microsoft data centre in West Texas. The project illustrates how major oil and gas producers are repositioning themselves to capture a new source of long-term energy demand.
For investors, the message is increasingly clear: the AI boom will not benefit only semiconductor manufacturers, cloud-computing companies and data-centre developers. It is also creating opportunities—and considerable infrastructure challenges—for natural-gas producers, utilities, turbine manufacturers and electricity-grid operators.
Watch the interview
Video: Fox Business via YouTube. Chevron CEO Mike Wirth discusses the company’s results, its Microsoft power agreement and the energy requirements of artificial intelligence.
Chevron moves into AI power generation
Chevron’s agreement with Microsoft represents one of the clearest examples yet of the convergence between the technology and energy sectors.
Under the twenty-year arrangement announced in June, Chevron plans to supply electricity to Microsoft’s expanding data-centre operations in West Texas. The project is expected to use natural gas to produce reliable power close to where it is consumed, reducing dependence on an already-constrained regional electricity grid.
Chevron has been developing the initiative with investment firm Engine No. 1 and energy-equipment manufacturer GE Vernova. The broader programme envisages several power facilities with a combined potential capacity of as much as four gigawatts.
That would be sufficient to serve multiple large computing campuses. More importantly, it would give Chevron a long-term contracted revenue stream linked to electricity rather than solely to fluctuating oil and gas prices.
In the interview, Wirth argued that artificial intelligence could influence the energy business for decades. Data centres require continuous electricity, making reliability just as important as price. Unlike many ordinary industrial consumers, cloud and AI facilities cannot easily tolerate interruptions.
Data-centre electricity demand could double by 2030
The scale of the emerging power challenge is considerable.
The International Energy Agency estimates that data centres consumed approximately 415 terawatt-hours of electricity in 2024, equivalent to around 1.5% of global consumption. Its base-case projection sees that figure more than doubling to approximately 945 terawatt-hours by 2030.
Data-centre electricity use could therefore grow by around 15% annually—more than four times faster than demand across the rest of the global economy. Electricity consumed by accelerated servers, the specialised machines principally associated with AI applications, is projected to rise by approximately 30% per year. The IEA’s detailed analysis also notes that cooling and other supporting infrastructure represent a substantial part of a data centre’s total power requirements.
The United States faces one of the most significant increases. According to the IEA, American data-centre electricity consumption could rise by around 130% between 2024 and 2030.
The problem is not simply the total amount of electricity required. Demand is concentrated in particular locations, placing enormous pressure on local generation and transmission networks.
Technology moves faster than the power industry
The difference between the development cycles of the technology and energy sectors creates a serious bottleneck.
A data centre can often be completed within two or three years. Large power stations, transmission lines and pipelines may require much longer because of planning procedures, environmental reviews, equipment shortages and permitting delays.
Wirth used the Fox Business interview to argue for faster US permitting reform. In Chevron’s view, the country possesses sufficient energy resources, but companies must be able to develop generation and infrastructure at the speed required by the AI economy.
This is the central warning for policymakers and investors: technology companies can announce billions of dollars in new computing capacity, but those investments cannot become fully productive unless reliable electricity is available when the facilities open.
The pressure is already visible in Texas. State authorities recently ordered additional scrutiny of proposed data-centre connections after projects under consideration produced electricity-demand requests vastly exceeding the system’s current peak load.
Why natural gas is returning to the centre of the debate
Renewable energy will remain an important part of the solution, particularly as technology companies pursue their emissions targets. However, solar and wind generation alone cannot always provide the continuous electricity required by hyperscale data centres.
Natural gas offers several advantages:
- Gas-fired plants can provide power regardless of weather conditions.
- They can be constructed more rapidly than nuclear facilities.
- Modern turbines can adjust production to balance variations in renewable generation.
- US gas supplies remain abundant, particularly in Texas and the Permian Basin.
- Generation located near a data centre can reduce dependence on congested transmission networks.
This does not remove the environmental debate. A sustained expansion of gas-fired generation could increase carbon emissions and complicate the climate commitments made by Microsoft and other technology companies.
Carbon capture, renewable natural gas and efficiency improvements may reduce the emissions footprint, but their economics and scalability remain uncertain. The AI power boom is consequently forcing technology groups to confront a difficult trade-off between rapid growth, electricity reliability and decarbonisation.
Chevron’s production strength supports the strategy
Chevron enters this new market from a position of substantial operational strength.
Wirth highlighted record US production and the continuing integration of Hess during the interview. The Hess acquisition expanded Chevron’s exposure to Guyana, one of the world’s most important emerging oil-producing regions, while also adding assets in the Bakken shale formation and the Gulf of Mexico.
The company reported that its second-quarter production increased approximately 20% from the previous year, reflecting both organic growth and the Hess combination. US output reached a record level of around 2.1 million barrels of oil equivalent per day.
Traditional oil and gas production therefore remains Chevron’s financial foundation. Electricity supply for AI represents an additional growth platform rather than a replacement for the company’s core business.
Long-term power agreements could nevertheless make Chevron’s cash flows more predictable. Oil prices can move dramatically in response to global supply, economic growth and geopolitics. A multi-decade electricity contract with an investment-grade technology company offers a very different risk profile.
AI creates a new group of energy winners
Chevron is not the only company benefiting from the acceleration of electricity demand.
Gas-turbine manufacturers such as GE Vernova and Siemens Energy have reported strong demand for generation equipment. Siemens Energy’s recent orders were supported by power projects associated with data centres and increased energy-security spending.
Utilities including American Electric Power, Dominion Energy and Constellation Energy are also attracting investor attention because of their exposure to expanding data-centre regions. Nuclear generation has regained strategic importance for the same reason: it can provide large quantities of continuous, low-carbon electricity.
The AI infrastructure trade is therefore broadening beyond Nvidia and the largest cloud-computing groups. Potential beneficiaries now include:
- Natural-gas producers and pipeline operators.
- Electricity utilities.
- Gas-turbine and grid-equipment manufacturers.
- Nuclear-power operators.
- Electrical engineering and cooling specialists.
- Renewable-energy and battery-storage developers.
However, not every project will generate attractive returns. Rapid investment could create excess capacity if AI adoption eventually falls short of current expectations. Developers must also contend with turbine shortages, rising construction costs, grid-connection delays and political resistance to the electricity and water consumption of data centres.
What it means for Chevron investors
Chevron’s power strategy gives the company exposure to a structural trend that could persist well beyond the current AI investment cycle.
The Microsoft agreement demonstrates that large technology companies are willing to enter exceptionally long contracts to secure electricity. That creates an opportunity for energy producers capable of combining fuel supply, power generation, project development and risk management.
For Chevron shareholders, the initiative could eventually produce three advantages:
- A new source of contracted long-term revenue.
- Greater demand for US natural gas.
- Diversification beyond traditional crude-oil production.
The financial importance of the programme will depend on construction costs, contractual returns and Chevron’s ability to deliver projects on schedule. Its contribution is still small compared with the group’s upstream oil and gas business.
Yet the strategic direction is important. Chevron is effectively arguing that abundant computing capacity requires abundant physical energy—and that the companies capable of supplying both reliably will occupy a critical position in the next phase of the AI economy.
The bottom line
The race to develop more powerful artificial-intelligence systems is becoming inseparable from the race to build electricity generation and transmission infrastructure.
Wirth’s warning is not that the world lacks energy resources. It is that power projects, grids and regulatory approvals may fail to keep pace with the speed of technology investment.
If that gap persists, electricity availability could become one of the principal constraints on AI expansion. If energy companies successfully close it, the data-centre boom could create a major new growth market for natural gas, utilities and infrastructure suppliers.
Chevron’s twenty-year agreement with Microsoft shows how the energy industry is beginning to respond—and why the next chapter of the AI investment boom may be written as much in power plants and transmission networks as in semiconductor factories.
