Artificial intelligence (AI) is moving from promise to proof, and the latest results from Meta, Microsoft, Amazon and Alphabet show how quickly the race is evolving. The broad message was clear – demand is strong, investment plans are still moving higher, and the biggest technology companies are starting to show where AI can support real financial growth.
But this is no longer just about who has the most exciting AI story. Investors are now asking a tougher question – can these companies turn huge spending into durable growth, stronger products and cash flows that justify the size of the bet?
This article isn’t personal advice. If you’re not sure an investment is right for you, seek advice. Investments and any income from them will rise and fall in value, so you could get back less than you invest. Ratios also shouldn’t be looked at on their own.
Investing in an individual company isn’t right for everyone because if that company fails, you could lose your whole investment. If you cannot afford this, investing in a single company might not be right for you. You should make sure you understand the companies you’re investing in and their specific risks. You should also make sure any shares you own are part of a diversified portfolio.
AI earnings are growing up
It’s hard to pick too many holes in the results from the four big AI investors. We saw double-digit earnings per share growth across the group, all either in line or ahead of expectations. This was a strong showing.
But this is no longer a market where “AI” is enough to get a free pass. The largest technology companies are spending at extraordinary levels to build the infrastructure needed for the next phase of growth. Investors are still willing to back that spending where the revenue link is clear, but the questions are getting harder.
Can this level of investment turn into durable growth? Can it support margins? And, eventually, can it produce the cash flows needed to justify the scale of the bet?
Investors are likely to reward companies that can grow while also showing clear evidence that today’s spending is creating tomorrow’s profits.
The AI spending boom is getting bigger
The second big message is that AI investment is not slowing down. If anything, the buildout is getting larger. Across Alphabet, Microsoft, Meta and Amazon, capital expenditure has already risen sharply over recent years, and forecasts point to another major step up in 2026 before growth slows in 2027 (though we think those numbers are too conservative).
The scale is striking. These four big spenders are expected to invest north of $670bn in 2026, up to more than $800bn in 2027. This reflects a simple reality. AI demand needs physical infrastructure. Data centres, chips, networking equipment, power, cooling and memory all have to be built before the revenue can be served.
With demand for AI computing power still ahead of available capacity, the companies with the deepest pockets have a strong reason to keep investing.
But the trade-off is cash flow. Heavy capital expenditure can make free cash flow look weaker in the short term, even when underlying demand is strong. That makes this spending cycle harder for investors to judge. The largest technology companies have the balance sheets to fund it, but the market will still want evidence that this is productive investment, not just an expensive arms race.
Investors want proof, not promises
The AI opportunity looks more real than it did a year ago, but investors won't reward spending for its own sake.
Not all AI investment is equal.
Spending on consumer AI features, new devices, social feeds or internal tools can still be valuable, but the returns are harder to model. Meta is a perfect example, where the benefits are being seen in the core advertising business, but the investment plans require belief that it’ll create new revenue streams.
The market is punishing that kind of higher-risk investment. We think this is where investors can still find some value, i.e. where the benefits aren’t as obvious as, for example, the cloud giants building more compute to rent out to third parties.
Cloud is where AI turns into revenue
Cloud remains the clearest link between AI investment and financial performance. The largest technology companies are not just building AI tools for themselves. They are selling access to the computing power, storage and software that other businesses need to build and run their own AI systems.
That makes cloud one of the most important areas to watch. It’s where AI demand can move from a long-term story into near-term revenue. Amazon, Alphabet and Microsoft are all benefiting from rising demand for AI-related capacity, and cloud revenue growth was a clear standout from recent results.
The numbers were strong. Microsoft Azure was up 39%, Google Cloud (GCP) rose 63%, and Amazon Web Services (which is growing from a larger base) at 28%.
The investment case for cloud is also easier to understand than some other areas of AI spending. If a company builds more cloud capacity and customers are waiting to use it, the revenue link is relatively direct. That does not remove the risk, but it makes the spending easier to justify than investment where the return depends on harder-to-measure improvements in engagement or future product adoption.
There is another advantage, too. The cloud giants can use the same infrastructure in two ways. They can rent it out to customers while also using it to improve their own products and services. That gives them a powerful position, but it also raises the bar.
The CPU is back in the spotlight
Taking a quick side-step, beyond the giant AI investors. One of the more interesting twists in the AI story from this earnings season is the return of the Central Processing Unit (CPU) into the limelight. Its sister chip, the Graphics Processing Unit (GPU), has taken most of the attention so far, and rightly so, but the use case for CPUs has expanded quickly as AI workloads become more complex.
Training huge models is only one part of the story. As AI moves further into inference, agentic workloads and real-world deployment, CPUs are playing a bigger role in orchestrating, managing and connecting these systems.
That shift has happened at remarkable speed. AMD has doubled its view of the long-term server CPU market in just a few months, from around $60bn by 2030 to roughly $120bn. That is a striking reset, especially from a company that should be extremely well placed to predict where chip demand is heading. It underlines just how fast this market is moving, and how quickly assumptions can change as the AI buildout broadens beyond the first wave of GPU-led demand.
For investors, the point is not that CPUs replace GPUs. They do not. It’s that the AI infrastructure opportunity is becoming wider and more complex.
The next phase is less about one type of chip doing all the work, and more about the full compute stack – CPUs, GPUs, networking, memory, software and custom silicon all working together. That makes the opportunity larger, but also harder to forecast in a market that’s moving faster than even some of its leading names expected.
The author holds shares in Meta and Microsoft.
This article is original Hargreaves Lansdown content, published by Hargreaves Lansdown. It was correct as at the date of publication, and our views may have changed since then. Unless otherwise stated estimates, including prospective yields, are a consensus of analyst forecasts provided by LSEG. These estimates are not a reliable indicator of future performance. Past performance is not a guide to the future. Investments rise and fall in value so investors could make a loss. Yields are variable and not guaranteed.
This article is not advice or a recommendation to buy, sell or hold any investment. No view is given on the present or future value or price of any investment, and investors should form their own view on any proposed investment.