SpaceX's AI Spending Spree Just Got a Specific Price Tag

The S-1 confirmed SpaceX burned $20 billion on AI in 2025. But the newer detail that's getting less attention: SpaceX announced a deal in April to acquire Cursor, an AI code-writing startup, for $60 billion, according to Scientific American.

Sixty billion dollars for a code-writing chatbot company. That's more than twice NASA's current annual budget.

It also happens to be roughly equal to what SpaceX expects to raise from its IPO, which according to TechCrunch is targeting around $75 billion at a reported valuation of $1.75 trillion, with a Nasdaq listing under the ticker SPCX expected later this year.

SpaceX plans to raise $75 billion from the public and immediately funnel most of it into AI acquisitions and orbital infrastructure. The company is using a rocket manufacturer's credibility to fund an AI conglomerate.

The Orbital Data Center Pitch Is Real — and Has a Date

Business Insider reported the core of SpaceX's long-term AI pitch: "AI compute satellites" deployed into sun-synchronous orbit, targeted for as early as 2028. The idea is to move AI infrastructure off Earth entirely.

The logic SpaceX is selling: terrestrial power grids can't keep up with AI demand. Space-based solar arrays generate more than five times the energy per unit area compared to ground-based systems. No weather. No permitting fights. No NIMBYs. Cooling happens through radiative heat dissipation — the vacuum of space does the work for free.

Business Insider noted that Google is pursuing the same concept, having launched its own related project late last year. When Google and SpaceX are chasing the same orbital compute thesis, the idea has moved past theoretical territory.

The xAI Problem

The AI division that SpaceX is betting its entire future on is losing money badly.

According to TechCrunch, SpaceX directed 60% of its 2025 capital spending — roughly $20 billion — to its AI division, which houses the Grok chatbot after absorbing Musk's xAI in February. That division lost billions last year. Its revenue grew only 22% in 2025.

For comparison, OpenAI and Anthropic are growing revenue at rates that dwarf that figure.

SpaceX has lost more than $37 billion since inception, per its own S-1. It lost $4.9 billion in 2025 alone on $18 billion in revenue, per Reuters. The company that Wall Street is pricing at $1.75 trillion is hemorrhaging cash at historic speed — and the AI bet is driving it.

Starship Is the Lynchpin — And It's Behind Schedule

Every part of this plan — orbital data centers, the NASA Artemis moon landings, Mars ambitions — depends on Starship working reliably at scale. According to Scientific American, Starship's development is delayed.

SpaceX is contracted to deliver a Starship-based Human Landing System for NASA's Artemis IV mission, targeted for 2028. Precursor Earth-orbit tests are scheduled for Artemis III next year. If Starship continues slipping, that entire timeline collapses — both the NASA contracts and the orbital AI infrastructure roadmap.

The Hill reported that SpaceX's near-to-mid-term growth plan depends heavily on both Starlink and Starship. Starlink is the one proven engine — it generated $11 billion in 2025, more than half of all company revenue. Everything else is a bet.

NASA Could Pay the Price

Scientific American raised a question that deserves more attention: what happens to NASA's Artemis program if SpaceX's AI pivot consumes all available resources?

Space historian Jordan Bimm of the University of Chicago framed the uncertainty: it's unclear whether "space is going to be the place where AI is used, or AI is going to be the means for us to do more in space."

SpaceX is also a major Pentagon contractor, supplying the military-grade Starshield version of Starlink and handling most U.S. defense launches. The company operates simultaneously as a defense contractor, an AI company, a satellite internet provider, and a rocket manufacturer — all with one man serving as CEO, CTO, AND chairman of the board.

TechCrunch noted that Musk will hold all three titles when SpaceX goes public. There is no independent check on Musk's decision-making built into the corporate structure. For a company seeking public investor money, that's a significant governance risk.

The Numbers Don't Hide

Most tech media is framing this as "SpaceX's exciting AI transformation."

The data tells a different story: SpaceX is raising $75 billion from retail and institutional investors to fund an AI spending spree that has already lost billions, acquiring companies at valuations that would make institutional investors pause, while its critical infrastructure (Starship) is delayed, and its CEO has zero board-level accountability.

The Cursor deal alone — $60 billion for a startup — is the kind of number that should trigger serious questions from IPO underwriters.

What This Means

If you're a taxpayer, you care because NASA's moon program hangs on Starship delivery schedules that are now competing with Musk's AI ambitions for attention and resources.

If you're considering investing, you're being asked to buy into a $1.75 trillion valuation on a company losing $4.9 billion a year, whose highest-growth segment (Starlink) is already mature, and whose AI division is growing at one-third the speed of its competitors.

The orbital data center idea might genuinely be the future. Or it might be the most expensive speculative pitch in IPO history. Right now, nobody — not SpaceX, not Wall Street, not NASA — actually knows which one it is.