Magnificent Seven's $700 billion 2026 capex plan faces overbuild risk; Meta CEO signals measured expansion caution.
Investors are growing concerned that the tech sector's massive spending on AI infrastructure may not generate sufficient returns. This anxiety echoes the digital infrastructure overbuild of the late 1990s that preceded the dot-com bubble.
The "Magnificent Seven" tech giants plan to spend more than $700 billion on artificial intelligence capital expenditures in 2026, a significant increase from approximately $400 billion in 2025. Throughout 2025, each announcement of increased AI-related capex boosted these companies' share prices. That dynamic has reversed: capex spending is now a major source of market tension, driven by investor concerns that returns may fall short of expectations.
Meta Platforms has become a focal point in this debate. CEO Mark Zuckerberg recently announced that Meta would launch a cloud business leasing excess compute capacity to external customers—a move that triggered an immediate stock rally. Meta has guided for 2026 capital expenditures of $125 billion to $145 billion, with most funding directed toward "additional data center costs to support future-year capacity."
The significance of this announcement lies in Zuckerberg's own earlier comments. In the third quarter of 2025, he suggested Meta would pursue compute leasing only if 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."
Despite the compute-leasing announcement, Meta's stock remained down approximately 9.5% year-to-date as of July 6, 2026.
Whether the industry faces genuine infrastructure overbuild remains uncertain. Space Exploration Technologies recently raised nearly $86 billion in its IPO, partly on the thesis of deploying a vast constellation of data center satellites in orbit—suggesting confidence in sustained compute demand. Additionally, rental prices for graphics processing units, including older models, are rising, indicating robust demand for compute resources.
Zuckerberg has characterized Meta's position not as outright overbuild but as building ahead of schedule. During the third-quarter 2025 earnings call, he noted that the worst-case scenario would be Meta bringing capacity online a few years early—incurring some asset depreciation, but eventually utilizing the compute.
The path forward remains contested. Investors should monitor both sides of the debate carefully; the narrative could resolve either direction. Critically, the hyperscalers have not yet deployed their full $700 billion in allocated 2026 capex and could revise spending guidance if conditions change. A pullback would likely alarm investors in individual "Magnificent Seven" stocks and could signal a major market red flag for the entire AI sector—a possibility that may already be partially reflected in these stocks' modest performance this year.