While many investors focus on chip makers and platform companies, AI infrastructure stocks could be the overlooked beneficiaries powering the buildout.
Every AI prompt requires computing power, and computing power requires energy to operate.
AI is becoming more efficient, but that’s only part of the story. More complex tasks like video generation, reasoning and agentic tasks can consume hundreds or even thousands of times more energy than a simple text query.
Due to the acceleration in adoption and rising query intensity, AI-focused energy consumption is in high demand.
That makes electricity a critical constraint on meeting the AI appetite.
As the International Energy Agency (IEA), which works with governments and industry on energy policy, puts it, “There is no AI without energy.”
Why does AI use so much energy?
While AI may appear to mimic human intelligence, at its core is a computer performing calculations.
Consider the difference between asking:
“What is the capital of France?”
And:
“What are the ten largest towns in Europe?”
The second requires far more computation: The AI must identify, gather and compare information; make assumptions about the request (what is meant by “largest”); and rank the final results before producing an answer.
More computation means more energy, the same way a triathlon will burn more energy than a sprint However, in this scenario, the event takes place in data centres (and other AI infrastructure) instead of a race course.
Global data centre spending is expected to exceed US$2 trillion over the next five years.
So, it isn’t a surprise to see that AI’s thirst for electricity has been rising rapidly.
In a 2025 published report titled Key Questions on Energy and AI, the IEA estimates electricity consumption from AI data centres surged 50% in 2025. It projects this to triple by 2030 in its base estimate. Elsewhere, estimates place data centres rising from around 3.7% of US power demand to nearly 12% by 2030.
Even when assuming improved efficiency, it still forecasts a significant increase in data centre electricity consumption.
Electricity is the fuel that powers AI and the need for that fuel is likely to be one of the major stories as the technology grows.
Why power is the near-term bottleneck
AI is arguably advancing faster than the infrastructure required to power it.
Some estimates calculate that total computing power used to train frontier AI models has doubled every five to six months. This echoes views held by the IEA, including the belief that AI has the potential to transform the energy sector.
There’s already evidence that a bottleneck is emerging.
According to the World Economic Forum, the investment into AI data centres and other AI infrastructure is accelerating faster than power grids were designed to accommodate. Meanwhile, critical materials used in energy production and data centre operation are seeing increased prices and supply constraints.
A lot of investor focus has been on chip makers.
But chips can be manufactured and deployed far faster than the electricity infrastructure required to power them. A computer chip can be produced in months. Building power plants, electricity grids and data centres can take years.
This makes AI infrastructure arguably the most critical, and overlooked, part of the artificial intelligence story.
Which stocks power the AI infrastructure expansion?
By this point the case for AI’s immense hunger for more energy should be evident. It’s a craving that could be much harder to satisfy than other pockets of the computing supply chain.
The Global X Artificial Intelligence Infrastructure ETF (ASX:AINF) is built around this premise and offers investors a single place to invest in a range of companies that provide the physical and operational backbone enabling AI’s expansion.
It’s a diverse list of companies spanning data centre operator Arista Networks (NYSE:ANET), French energy technology business Schneider Electric (ENXTPA:SU), copper miner Antofagasta PLC (LSE:ANTO), power management company Eaton Corporation (NYSE:ETN), as well as energy giant Seimens Energy (XTRA:ENR), to name a few of its 30 holdings.
What does it mean for investors?
Artificial intelligence is one of the core themes facing global markets today.
With more use cases of the technology constantly being developed, there is little doubt that demand for artificial intelligence will continue into the future.
However, while many focus on chip manufacturers and AI companies providing the models we use, there is a whole supply chain required to make the technology work, and that is where AI infrastructure stocks like those held by the Global X Artificial Intelligence Infrastructure ETF come into play.
While the frontier AI companies are providing the platform, AI infrastructure stocks (like data centre providers, energy grid companies and even mining companies) are providing the critical backbone.
Frequently asked questions
How much energy do data centres use?
The exact numbers are unclear but IEA estimates currently place data centres at around 1.5% of global electricity consumption in 2025. This is estimated to grow to as much as 3% by 2030.
Why is AI straining the grid?
Data centres use enormous amounts of electricity, not just to operate the servers housed inside them but to also cool them. Currently, the investment in building new data centres is outpacing investment in the electricity grids required to allow them to remain operational.
What is an AI infrastructure ETF?
An AI infrastructure ETF is a portfolio of AI infrastructure stocks or companies that make up the AI Infrastructure supply chain. This includes data centre providers, energy companies, energy management and product manufacturers and even miners who mine critical materials like copper and uranium which are needed in the supply of energy or the operation of data centres.
A power plant waking up and supplying energy is the first step in going from an AI prompt to an AI output.
That physical layer is where the most followed Simply Wall St community Narrative for the Global X Artificial Intelligence Infrastructure ETF sees an investment opportunity. Rather than owning the crowded chip and software names, this ETF offers a different exposure to the constraints that could benefit amid growing AI demand.
Without the power-providing infrastructure, there is no way for the AI model to perform the calculations required. As such, it’s arguably one of the most critical components underpinning civilisation’s ability to leverage the powerful technology.
So, while AI might be framed as a technology story, it is, first and foremost, arguably an energy story.
Disclaimer
This article was written independently by the author, without issuer input or approval. Global X Management (AUS) Limited has a marketing services agreement with Simply Wall St. Details on compensation and other important information can be found in the disclosure and disclaimer at the end of this narrative.
Have feedback on this article? Concerned about the content? Get in touch with us directly. Alternatively, email editorial-team@simplywallst.com
Simply Wall St analyst Andrew Legget and Simply Wall St have no position in any of the companies mentioned. This article is general in nature. We provide commentary based on historical data and analyst forecasts only using an unbiased methodology and our articles are not intended to be financial advice. It does not constitute a recommendation to buy or sell any stock and does not take account of your objectives, or your financial situation. We aim to bring you long-term focused analysis driven by fundamental data. Note that our analysis may not factor in the latest price-sensitive company announcements or qualitative material.