Is the "Magnificent Seven's" Plan to Spend $700 Billion on AI Capex in 2026 Going to Lead to an Overbuild? Meta's CEO Mark Zuckerberg May Have Just Revealed the Answer.

The "Magnificent Seven" plan to spend more than $700 billion on artificial intelligence capital expenditures this year, a big step up from the $400 billion or so the group spent in 2025.

In 2025, whenever hyperscalers announced plans to increase their AI-related capex, their stocks surged. But now, that spending has become a major point of contention in the market, primarily because investors are worried that the returns on these massive investments may not live up to the hype.

In particular, investors are worried that hyperscalers may overbuild AI infrastructure. Meta Platforms (NASDAQ: META) CEO Mark Zuckerberg may have just given us a big hint about how valid those concerns might be.

Image source: Getty Images.

Meta's new cloud infrastructure plan could be a tell

Recently, Meta announced it is launching a new cloud business that will lease its excess compute capacity to external customers. Shares popped on the news, as it could lead to immediate revenue from the company's new data center builds, which investors are already clamoring to see, given the size of Meta's capex.

Meta has guided for capital expenditures of $125 billion to $145 billion this year, most of which will cover "additional data center costs to support future-year capacity."

The announcement is big news in the AI narrative because back in the third quarter of 2025, Zuckerberg implied that his company wouldn't become a supplier of compute unless it overbuilt AI infrastructure:

Now, I mean, it's of course possible to overshoot that, right? And if we do, I mean, this is what I mentioned in my comments, then we see that there's just a lot of demand for other new things that we build internally, externally. Like, almost every week, people come to us from outside the company asking us to stand up an API service or asking if we have different compute that they could get from us. And we haven't done that yet, but obviously, if you got to a point where you overbuilt, you could have that as an option.

Now, it's not a total surprise, as Zuckerberg has been hinting that Meta might begin leasing compute, and the stock has struggled this year. Even after the rally on the cloud announcement, the stock was still down about 9.5% year to date as of July 6.

Does this signal a massive overbuild?

As with everything else in AI, it's hard to provide a definitive answer on whether we are at the beginning of a massive overbuild in AI infrastructure. After all, consider that Space Exploration Technologies recently raised nearly $86 billion in its massive IPO, partly on the thesis that it will deploy an enormous constellation of data center satellites in orbit.

Zuckerberg also does not necessarily view the current situation as an outright infrastructure overbuild; rather, it is that the company has gotten ahead of schedule in building what it will require. On the company's third-quarter 2025 earnings call, he also said that the worst-case scenario is that Meta has built some of its AI data center capacity a few years in advance. In this scenario, while those assets would experience some loss and depreciation, the company will eventually utilize the compute.

Additionally, rental prices for most graphics processing units (GPUs), even older models, appear to be on the rise, suggesting that demand for compute remains strong.

All that said, investors should continue to weigh the evidence carefully on both sides of the debate, and understand that the narrative could break in either direction. Furthermore, the hyperscalers have not yet spent the full $700 billion that they've allocated to capital expenditures this year. They could easily revise their AI capex guidelines should conditions require it.

If there is a pullback in spending, while investors in individual "Magnificent Seven" stocks may feel relieved, the market could view it as a major red flag for the entire AI trade.

Perhaps this scenario has been somewhat priced into these stocks, given the group's struggles thus far this year, but it's a risk investors need to be cognizant of, and Zuckerberg may have given the market a glimpse of what's to come.

Bram Berkowitz has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Meta Platforms. The Motley Fool has a disclosure policy.