Why the AI boom could extend the natural gas era

Artificial intelligence is often viewed as a driver of higher electricity use and, by extension, faster decarbonization. However, one of its most immediate effects may be the opposite of what many assume. The rapid expansion of AI infrastructure is increasing demand for reliable power, and this reality could reinforce the role of natural gas and other dispatchable energy sources for many years.

Investors focused on semiconductor and software valuations may be overlooking a fundamental constraint. AI runs on electricity, and power systems operate within physical and economic limits.

AI Is Driving a New Wave of Power Demand

The energy sector spent most of the past decade dealing with slow growth in electricity loads. That is now changing, in a way reminiscent of the sharp rise in oil demand — and then prices — in the early 2000s.

Training large language models and running advanced AI systems require massive computing resources. Hyperscale data centers are expanding rapidly, with developers requesting gigawatt-scale grid connections from utilities. In several regions, electricity demand forecasts have been revised upward after years of flat expectations.

The significance of this shift is that AI workloads generate continuous, high-density demand rather than intermittent usage. Data centers cannot simply shut down when power supply becomes constrained. Reliability becomes critical.

Reliability Needs Are Changing the Generation Mix

Wind and solar capacity continues to expand, but intermittent generation alone cannot meet the firm capacity needs of AI infrastructure without large-scale storage or backup generation.

Battery storage is improving, but long-duration storage remains expensive at scale. Nuclear projects face long development timelines and regulatory complexity. Transmission expansion is also lagging demand growth in many regions.

These constraints make dispatchable energy sources essential. Natural gas plants can ramp output quickly, run continuously, and be deployed faster than many alternatives. As a result, gas-fired generation is increasingly viewed as a practical solution for supporting AI-driven load growth.

This does not eliminate the role of renewables. In many markets, new renewable capacity is paired with gas generation to maintain grid stability. The key point is that AI-driven electrification of demand is likely to increase fossil fuel use in the near term.

Natural Gas May Be One of the Biggest AI Winners

Several factors support natural gas as a near-term beneficiary.

Construction timelines favor gas plants when demand rises quickly. Existing pipeline infrastructure lowers expansion barriers. For data center operators, reliability usually outweighs ideological preferences, since outages are extremely costly.

Utilities are also revising resource plans as load forecasts rise. This shift could drive higher investment in transmission networks, grid upgrades, and flexible generation assets.

The Decarbonization Story Is More Complex

A common narrative holds that AI accelerates the shift away from fossil fuels because it increases electrification. The reality is more nuanced.

If electricity demand grows faster than low-carbon capacity, fossil generation may rise in absolute terms even if renewables gain market share. Total emissions could increase while carbon intensity declines as cleaner sources take a larger share of supply.

Energy systems ultimately evolve based on engineering and economics, not only policy goals or market narratives.

What Investors May Be Missing

AI is often discussed as a technology story, but it is also an infrastructure story.

Rising power demand could benefit utilities investing in transmission and generation capacity. Natural gas producers and midstream infrastructure companies may see structural demand support from higher power-sector consumption. Suppliers tied to grid reliability equipment and gas turbines may also benefit.

Over the longer term, advances in nuclear, storage, or efficiency could change the trajectory. For now, the immediate response to a sharp rise in electricity demand is likely to rely on technologies that can be deployed quickly and reliably.

AI may reshape the economy in profound ways. One of its most underappreciated effects is that it could extend the importance of natural gas while the world builds the energy backbone needed for the next generation of computing.