Everywhere you look right now, the obsession with artificial intelligence is impossible to miss. From chipmakers printing money to startups reinventing themselves overnight as “AI companies,” it feels like this theme has completely taken over the markets.
In fact, who am I kidding, it doesn’t just feel like it has. It all but certainly has.
But AI isn’t just about Nvidia, semiconductors, or the next flashy LLM. It’s a structural shift that’s quietly touching every single corner of the market, from utilities racing to expand power grids to REITs rethinking data center capacity, to software firms embedding automation into their business models just to stay competitive.
Many have asked me to highlight some potential AI ETFs they can take advantage of. And you can probably see why they’d be confused when I started mentioning ETFs that contain engineering firms and utility companies.
This is because AI ETFs come in many forms.
We can see pure-plays chasing the hardware and software backbone, engineering firms building out data centers, or grid construction companies that are supplying transmission lines.
In this issue of ETF Insights, I’m going to give you some of the best AI ETFs that are around today.
However, I’m not going to point you to the blatantly obvious ones like CHPS, SOXX, etc.
Instead, I’m going to show you how you can dig deeper to benefit from some of the areas of the market that will no doubt benefit from AI, but haven’t quite got the spotlight on them yet.
Why Utility ETFs Could Be the Stealth AI Winners of the Decade
I find this is arguably the most fascinating area of the market regarding AI. After all, we’re talking about slow-growing, boring utilities.
But, I will show you a chart of Capital Power (TSE:CPX) below.
The company has gone up 121% in just two years. Prior to 2023, it took the company almost eight years to increase by ~120%.
Why did Capital Power do so well? Well, it will be one of the primary power sources for data center providers in Alberta.
AI’s demand for electricity is rewriting the outlook for utilities. Data centres now use far more energy than traditional computing. Because every company on the planet wants a piece of AI, demand is skyrocketing.
Utilities are a true “picks and shovels” play. No matter how profitable a piece of AI software is, utilities are an essential upstream element. The piece of AI software could be losing millions of dollars a day. The tech still needs the energy to operate, no matter what.
However, not all utilities are the same. Each company can be affected differently by external factors, such as the increased focus on AI we are seeing. Traditional, regulated utilities (the ones that earn a fixed rate of return approved by regulators) won’t see the same upside.
Their earnings are designed to be stable, not explosive. We can think of a company like Fortis. This doesn’t necessarily mean that Fortis is a bad company. This simply means that, viewed through a goal of finding connections to the AI industry, Fortis is unlikely to be a true AI winner.
The real winners are likely to be independent power producers and grid builders. These companies can respond faster to market demand, negotiate private contracts with data center operators, and expand capacity without the same regulatory constraints.
In other words, while regulated utilities will keep the lights on, it’s the independent players building and selling that power who stand to profit the most from AI’s energy boom.
So, what are the best ETFs for this?
In the US, I tend to lean towards the Utilities Select Sector SPDR Fund (XLU).
The fund does have a lot of regulated utilities, but it also has a lot of individual power producers. Some of the utility companies that are expected to be the primary beneficiaries of AI power demand (NextEra Energy, Constellation Energy) are the top holdings in the fund.
Canada is a bit more difficult. This is because a lot of our utility companies are regulated utilities, and as such, many of our ETFs will be filled with them.
However, there is a fairly unique one in the fact that it contains a lot of independent power producers and even renewable options, and that is the BMO Equal Weight Utilities Index ETF (ZUT.TO).
The two largest holdings in this fund are independent power producers in Capital Power and Transalta. I feel this is one of the better utility ETFs in the country, regardless of association to the AI sector.
AI’s Role in Boosting Transportation
AI is quietly reshaping how freight moves. Many transportation and logistics companies now lean on algorithms to trim costs and manage volatility in freight volumes.
Even companies like Waste Connections and Waste Management are piloting AI utilization in sorting facilities. Instead of humans identifying where particular pieces of trash go on conveyors, AI is doing it. The company mentions this could lead to a 50% reduction in labor costs.
This particular sector is an area that nobody is looking at right now in terms of improvements via AI.
Studies show that AI in logistics improves efficiency and reduces idle time, especially in last‑mile delivery. When margins are tight (which is the case with many of these trucking and logistics companies), a slight improvement in terms of efficiency can really boost operating income.
Autonomous vehicles and predictive maintenance are the next elements here. AI is already very good at monitoring fleet health, cutting downtime and thus extending the asset life of the vehicles. Look no further than a company like Uber, which is utilizing this already.
This is more of a long-term bullish outlook regarding AI. Why? Scaling it across fleets takes a lot of capital and regulatory approval. Both of those are slow-moving variables. However, if they do end up working out, there could be some big tailwinds there for a long time.
In terms of ETFs to grab this type of exposure, unfortunately, there aren’t any in Canada. However, there is a solid one in the US, and it is the iShares US Transportation ETF (IYT).
The fund has extensive exposure to Uber and Union Pacific, making up around 38% of the portfolio. However, these are two high-quality names, both of which should be able to benefit from AI.
You’re unlikely to see large-scale returns from this one like you would with pure-play AI software companies. However, this is a sneaky area of the market in regards to AI, one that could have tailwinds for many years.
Data Centres Are the New Oil Fields
Every new surge of AI demand pushes deeper into physical infrastructure. The servers running large models need power, cooling, space, chips. All this means increased demand for data centres.
Many ETFs in this area give investors a direct line to that build‑out. They hold companies that own or lease digital infrastructure. Think of things like towers, fibre networks, and hyperscale server farms.
The revenue model is simple enough for anyone to understand: long‑term leases tied to data usage and power capacity.
Demand for AI compute infrastructure rises as firms train and deploy larger models. Operators like Equinix and Digital Realty keep adding capacity, often at double‑digit megawatt growth rates.
This industry seems promising. Similar to a residential REIT, you buy it, benefit from the rents, and get paid the vast majority of the REIT’s profits.
However, I do have concerns with this industry. They aren’t necessarily alarm bells. If they were alarm bells I wouldn’t be highlighting this as an option. Still, they’re definitely a “keep an eye on this” situation.
The main issue would be the fact that many of these data center REITs’ revenue streams come from the hyperscalers. And what are the hyperscalers doing right now? They’re building out extensive data centers. If capacity catches up with demand, we could see pressure on rents.
I don’t see it happening any time soon, but it’s a possibility.
There is pretty much one ETF that handles this area, and it is the Pacer Data & Infrastructure RE (SRVR).
The fund contains many of the top companies in the industry such as Equinix, Digital Realty, American Tower, etc.
This potential path is likely more appealing to income-focused investors, as a lot of these companies are on the higher-yielding end, with the ETF yielding nearly 3%.
This ETF has lagged the market extensively over the last while. We do have to remember that a lot of the companies inside of this fund are real estate-based.
Real estate has had a really rough go over the last while here with rising rates. However, declines south of the border could be a tailwind.
AI in Manufacturing
Manufacturing used to move at a human pace. Now, algorithms schedule shifts, sensors track output, and robots handle precision work without taking 3 breaks per day.
The shift in manufacturing is no longer theoretical. This change is visible across smart factories, predictive maintenance systems, and production lines.
If you remember earlier, I mentioned how the garbage disposal companies are using AI to sort garbage. Something that could trim labor costs by 50%.
Predictive algorithms are identifying equipment failures before they occur, drastically reducing downtime. Machine systems are inspecting products with absurd precision, improving quality control beyond what humans can achieve.
There are artificial intelligence drones that exist to help farmers spray crops only where weeds exist, saving them huge amounts of time and money.
Predictive maintenance has the potential to impact a wide range of companies. If you think about it, if AI can detect maintenance issues before they occur, it can stop or at least reduce production halts.
These types of ETFs offer indirect AI exposure without relying on semiconductor momentum. They also don’t rely on largely unproven AI-related business models. A lot of these companies are already seeing benefits from AI-related improvements.
In terms of ETFs for exposure, the only quality one we have in Canada is the iShares S&P Global Industrials Index ETF (XGI.TO).
Not only does it track Canadian and US industrial names, but global names as well.
Keep in mind, this is a Canadian-listed fund that simply owns a US-listed ETF, the iShares Global Industrials ETF (EXI). So, this can cause some taxation issues on the dividend, but nothing outrageous. In addition to this, it is Canadian hedged.
On the US side, it is hard to argue against the Vanguard Industrials Index Fund ETF (VIS).
You’re getting exposure to around 400 companies in the automation/manufacturing/equipment industries, like GE Aerospace, Caterpillar, Uber, Boeing, John Deere, Waste Management, etc.
The Case for Energy ETFs
AI isn’t just a story about chips and data. It’s also a story about the raw materials that make it all possible.
Training and running large models requires massive power inputs, and that energy demand keeps climbing. The International Energy Agency expects global electricity use from data centres, AI, and crypto to more than double from 460 terawatt-hours to over 900 by 2026. Imagine what that looks like by 2030.
That kind of demand will reshape the entire energy landscape. More computing means more cooling, more transmission, more backup generation, and more pressure on the grid.
And while most people think “oil and gas” when they hear energy, it goes much further than that. Renewables, nuclear, storage, and the materials that feed them. Uranium producers like Cameco, for example, could quietly become key beneficiaries as demand for steady, more carbon-free power grows.
Of course, energy remains cyclical, and commodity prices swing wildly. Regulatory and environmental pressures may hold back new projects or drive up costs. However, I do feel governing bodies are going to start to become a lot more lenient when it comes to regulatory approvals. After all, who wants to be known as the government that held their country back from making rapid AI advancements due to political hurdles?
The key metric to watch is energy intensity per AI model. If efficiency gains lag behind computing growth, then energy-linked ETFs, from broad energy funds to nuclear and utility plays could remain among the most relevant ways to invest in AI’s expansion long after the hype around chips fades.
Two interesting funds to look at here are Global X Uranium ETF (URA) and iShares Global Energy ETF (IXC).
URA provides exposure to uranium miners and nuclear technology companies. A space that could quietly become one of the biggest long-term beneficiaries of AI-driven energy demand. If global data center expansion accelerates, stable, carbon-free power will become indispensable, and that is what nuclear offers.
One caveat for URA, however, is that it will be especially volatile. If you’re holding this one long-term, be well prepared for 40%-50% swings in price to both the upside and downside over short periods of time.
IXC, on the other hand, holds global oil and gas majors like Exxon, Chevron, and Shell, the firms that will still supply the backbone of global energy infrastructure even as renewables scale. It offers broader exposure to the energy supply chain, from production to refining and distribution, and can serve as a hedge against rising electricity demand.
Overall, don’t get tunnel vision on semiconductor plays when it comes to AI
AI’s impact on markets goes far beyond tech stocks. However, this is where the vast majority of eyeballs are at this point in time.
The beauty of this shift is that it opens the door to opportunities in areas that investors rarely associate with innovation. Utilities, energy production, transportation networks, and industrial firms are all becoming part of the AI ecosystem.
Not because they’re building chatbots or models, but because they supply the physical backbone that those systems depend on.
The best part about these alternative routes, such as utilities, manufacturing, industrials etc., is that if AI ends up being not as promising as predicted, they’ll still continue on with their operations. Pure-play AI companies trading at nosebleed valuations? They’re likely to face substantial drawdowns.
Sure, we won’t hit gigantic home runs with these options. However, as you continue to invest, you’ll find how beneficial it is to hit consistent singles and doubles over trying to hit it out of the park all the time.
The market will keep chasing the next hot ticker in AI. You don’t have to. Sometimes the smartest way to play a revolution is to own the companies that are keeping the lights on for the new big thing.
