top of page
Nov 10
3 min read

COP 31: Artificial Intelligence Is Miring Green Energy Transition


November 2026. The halls of COP31 are brightly lit, and national delegates continue to repeat the official language of “accelerating energy transition” and “phasing out fossil fuels”. Outside the conference venue, however, a harder reality is unfolding in parallel: artificial intelligence demand for electricity is expanding at a pace that far outstrips the rhythm of climate negotiations.

 

“The end of computing power is electricity.” This statement, voiced in recent years by both Elon Musk and Jensen Huang, is being confirmed with increasing frequency. Growing attention is also being paid to the energy sources behind every kilowatt-hour. Looking back to around 2020, ESG was still the dominant narrative: Microsoft pledged to become a carbon-negative company, BlackRock Chairman Larry Fink publicly warned that the climate crisis would determine whether companies could survive in the long run, and the World Economic Forum in Davos placed climate at the top of its agenda. These commitments did not vanish overnight. Over the past few years, they have been gradually diluted by more pressing realities—energy security, economic development, and now the race for AI computing power.

 

Just last year, after COP30 concluded in Brazil, many were startled to discover that the once-shared goal was being quietly abandoned. COP30 produced no concrete resolution on phasing out fossil fuels. Some oil-producing countries argued that they should be allowed to continue using fossil fuels to develop their economies. Against the backdrop of AI’s rapid expansion, this position has become more persuasive: AI’s hunger for electricity is restoring strategic value to stable, dispatchable fossil fuels.

 

This value is not limited to power generation; it is also penetrating oil extraction itself. Take the Permian Basin, which accounts for roughly half of U.S. oil production. Because the oil is trapped in shale, companies have long relied on hydraulic fracturing (fracking): high-pressure mixtures of water, sand, and chemicals are injected deep underground to fracture the rock and release the oil. These laterally drilled wells, which can extend for miles, typically decline very rapidly—causing production often falls by as much as 70 percent in the first year.

 

To address this problem, Chevron developed its own AI tools to predict which underground sections will yield the most oil and which rock layers will make drilling more difficult. Similar examples are widespread. Schlumberger has deployed automated drilling systems and uses AI even in offshore operations for seismic surveys and well planning. The result of these technological advances is that, over the past decade, U.S. daily oil production has risen by 60 percent while the number of workers has fallen by 40 percent. Goldman Sachs has projected that if oil companies widely adopt AI, they could accelerate project timelines, significantly cut capital expenditure, and reduce the price of crude by as much as $11 per barrel. This is not merely a decline in cost; it is a reshaping of competitiveness.

 

When the computing race elevates electricity to a strategic priority, stable and dispatchable fossil fuels regain their rationale. At the same time, when AI is used to raise the efficiency of oil extraction and lower its costs, the “cost advantage” that the green transition once counted on is also eroded. BlackRock Chairman Larry Fink has recently been blunt about the power bottleneck: “We have enough electricity in the United States, especially through natural gas, but we can’t distribute it properly… If we don’t do that, we are not going to succeed in AI. AI is essentially a bunch of electrons. You need power to create those electrons.”

 

The path from the ESG consensus of 2020 to today’s pressures of computing power and electricity is not a sudden turn. It is the result of being gradually squeezed by more urgent constraints. Artificial intelligence is both a massive consumer of energy and an enhancer of fossil-fuel efficiency. Under this dual role, truly pulling the green energy transition out of the quagmire is likely to prove far more difficult than people imagined just a few years ago.


Author: Mr. Parco Siu, Project Assistant of ICSD
website-logo.png
International Chamber of Sustainable Development
CEP-transparent background_edited.png
CEM_logo_transparent_edited.png
  • Facebook
  • LinkedIn
  • Youtube

© 2026 International Chamber of Sustainable Development Limited | All Rights Reserved

bottom of page