SpaceX has spent two decades building its reputation as a rocket company, but by its own CEO’s account, that’s no longer the biggest part of the business — AI is. Our Elon Musk Business Empire guide covers the broader context of how SpaceX fits into Musk’s wider company network.
What Musk Actually Told Employees
In a 30-minute all-hands video released on X, Musk told SpaceX staff plainly: “Probably our AI revenue — not probably, definitely — our AI revenue will exceed all other SpaceX revenue probably in September.” This is a genuinely striking claim for a company whose public identity has been built almost entirely around rockets and satellite internet. Yahoo Finance’s coverage of the announcement reported the statement alongside SpaceX’s underlying financial breakdown.
The Numbers Behind the Claim
SpaceX’s second-quarter 2026 report showed total revenue of $7.81 billion, up 92% year-over-year — a genuinely dramatic growth rate for a company already operating at massive scale. Of that total, the AI segment contributed approximately $2.56 billion, while Connectivity (Starlink) and Space (launch services) combined brought in $5.25 billion. For AI to overtake the rest of the business by September, as Musk claims, that segment would need to more than double in a matter of weeks — an aggressive but not impossible trajectory given the pace of recent contract signings.
Where This AI Revenue Is Actually Coming From
This growth is being driven by new large-scale compute contracts, reportedly including deals with Anthropic and Alphabet’s Google. This reflects the practical outcome of SpaceX’s February 2026 all-stock acquisition of xAI, which formally merged the rocket company with Musk’s AI venture, combining their infrastructure and business development efforts under one corporate roof.
Analyst estimates back up the scale of this shift — Goldman Sachs modeled roughly $15.6 billion in AI revenue for SpaceX across all of 2026 following the two major compute deals. That’s a substantial forecast for a business line that barely existed as a distinct, reported category a year earlier. Our guide to smart investment strategies touches on how quickly business models can shift when a company leans into a high-growth adjacent market.
The Power Constraint Nobody Can Ignore
Musk has also set a goal of reaching 10 gigawatts of AI compute capacity by the end of 2027 — an enormous target that runs directly into one of the AI industry’s most persistent bottlenecks: electricity. SpaceX’s current data centers, known as Colossus 1 and Colossus 2, have drawn real pushback for relying on methane-producing gas turbines to meet power demand, a tension between rapid AI infrastructure buildout and environmental scrutiny that isn’t unique to SpaceX but is particularly visible given the company’s public profile.
Why This Matters Beyond SpaceX Itself
- It signals how thoroughly AI compute demand is reshaping companies that weren’t originally built around AI at all
- It shows how quickly a major acquisition (the xAI merger) can shift a company’s entire revenue mix within months
- It highlights a growing industry-wide tension between AI’s massive power needs and environmental constraints
- It suggests infrastructure and compute contracts, not just consumer AI products, are becoming a major profit center across the industry
A Reasonable Dose of Skepticism
It’s worth noting Musk has a documented history of optimistic, sometimes premature predictions about his companies’ timelines — from Tesla’s self-driving promises to earlier AGI predictions that didn’t materialize on schedule. Whether AI revenue actually overtakes the rest of SpaceX’s business by September specifically, or simply grows to be a much larger share slightly later, the underlying trend — AI compute becoming a major and rapidly growing SpaceX revenue line — is well-supported by the reported Q2 numbers themselves, independent of the exact timeline claim.
How This Compares to Other AI Infrastructure Plays
SpaceX’s approach to AI infrastructure differs from pure-play AI companies in a meaningful way: it’s leveraging existing aerospace-grade engineering discipline, capital access, and large-scale project management experience rather than building compute infrastructure expertise from scratch. This gives the combined SpaceX-xAI entity a genuinely different competitive profile than AI labs that lack that kind of heavy-industry operational background, though it also means the company is managing an unusually wide span of technical domains simultaneously — rockets, satellites, and AI compute all under one roof.
What This Signals About the Broader AI Infrastructure Race
SpaceX isn’t alone in discovering that AI compute demand rewards companies with existing large-scale infrastructure and capital deployment experience. Traditional cloud providers, energy companies, and even some industrial firms have found themselves newly relevant to the AI conversation simply because building and powering data centers at the scale modern AI models require draws on exactly the kind of engineering and capital-raising experience those companies already had, well before the current AI boom made it valuable in this specific new context.
What This Means for Investors and Business Watchers
For anyone tracking SpaceX given its 2026 IPO, this shift matters directly — a company increasingly earning its growth from AI compute contracts carries a different risk and valuation profile than a pure aerospace and satellite business, tied more closely to the broader AI infrastructure buildout than to launch cadence or satellite subscriber counts. Our buying vs leasing decision framework covers a much smaller-scale version of the same principle — understanding what’s actually driving a asset’s value before committing capital to it.
The Bottom Line
SpaceX’s own CEO says AI revenue will soon exceed its combined rocket and satellite internet business — a claim backed by real, reported Q2 2026 financials showing 92% year-over-year revenue growth and a rapidly scaling AI segment. Whether the exact September timeline holds, the broader shift is real and significant: one of the world’s most recognizable aerospace companies is, by revenue mix, becoming an AI infrastructure company as much as a rocket one.
