SpaceX is making a massive wager that artificial intelligence, not rockets, will define its future. The company spent $12.7 billion on AI in 2025, more than three times what it invested in its core space business, according to its recent Securities and Exchange Commission filing ahead of its June 12 initial public offering (IPO). That aggressive spending came with a $6.3 billion operating loss in the AI segment alone, dwarfing the $657 million loss from its rocket operations.

For investors eyeing SpaceX's public debut, the numbers raise an immediate question: Is Elon Musk's company pivoting away from space exploration toward artificial intelligence infrastructure? The answer is more nuanced. SpaceX is betting that the two businesses will eventually merge, with space-based data centers powering the next generation of AI models including Grok 5, the company's frontier language model.

Where Is All That AI Money Going?

The bulk of SpaceX's AI spending is flowing into two massive data centers called Colossus and Colossus II, which train the company's next-generation frontier models. These facilities are extraordinarily expensive to operate because they consume vast amounts of electricity. SpaceX has identified a solution that sounds like science fiction but is grounded in practical engineering: move the data centers into orbit.

The company explained its vision in its SEC filing: "The logical path forward is to move power-intensive AI workloads into orbit, where solar energy is near-constant and uninterrupted". This would involve large satellite constellations linked together and equipped with AI processing capabilities to function as orbital data centers. By placing these facilities in space, SpaceX could sidestep two major obstacles facing terrestrial AI infrastructure: the enormous energy demands of training large language models and growing community backlash against building data centers on Earth.

How Is SpaceX Offsetting These Massive Losses?

SpaceX is not sitting idle while losses mount. The company recently signed a cloud services agreement with Anthropic, the AI company behind Claude, that will generate $1.2 billion per month in revenue through mid-2029. This deal gives Anthropic access to SpaceX's Colossus data centers and represents one of the largest commercial AI infrastructure contracts announced to date.

Additionally, SpaceX estimates its total addressable market in AI, including enterprise applications, AI infrastructure, consumer subscriptions, and advertising, at $26.5 trillion. The company is betting that early investments in data center technology and AI model development will position it to capture a meaningful share of this enormous market as demand for AI computing power accelerates.

What Are the Key Risks and Uncertainties?

• Unproven Technology: Orbital data centers remain largely theoretical. While Alphabet and Amazon have acknowledged the concept's feasibility, Amazon founder Jeff Bezos cautioned that timelines of just a few years "are probably a little ambitious".
• Rising Costs: SpaceX's AI spending is accelerating, not slowing. Capital expenditures on AI reached $7.7 billion in the first quarter of 2026 alone, suggesting annual spending could exceed $30 billion if the trend continues.
• Distraction from Core Mission: SpaceX's stated goal is to transform rocket launches into "airline-like operations" and send humans to Mars. The massive diversion of resources to AI infrastructure raises questions about whether the company can pursue both ambitions simultaneously.

Why Does This Matter for the Broader AI Landscape?

SpaceX's pivot reflects a fundamental shift in how technology companies view the AI bottleneck. For years, the limiting factor in AI development was algorithmic innovation. Today, it is raw computing power and the energy to run it. Companies like OpenAI, Anthropic, and Google are all racing to secure access to massive amounts of GPU (graphics processing unit) capacity, which is why SpaceX's willingness to spend billions on infrastructure is strategically significant.

The company's merger with xAI in February 2026 signals Musk's intention to vertically integrate AI development, from model training to infrastructure to deployment. Tesla is already exploring autonomous systems and humanoid robotics, and there is speculation that SpaceX and Tesla could merge as soon as next year, creating a conglomerate that spans AI, space, energy, and transportation.

What Should Potential SpaceX Investors Know?

SpaceX's IPO will force public shareholders to make a bet on Musk's vision of the future. The company is not hiding its AI ambitions or its losses. In fact, SpaceX explicitly acknowledges in its SEC filing that AI costs are rising and will likely continue to do so. This is a company asking investors to fund a multi-year, multi-billion-dollar experiment in space-based AI infrastructure while simultaneously pursuing its original mission of revolutionizing space travel.

For risk-tolerant investors who believe in Musk's long-term vision, SpaceX's AI spending could prove to be a prescient bet on the future of computing. For more conservative investors, the company's willingness to lose billions annually on an unproven technology may be a red flag. As one analyst noted, "The amount of worry investors should have is probably proportional to the amount of risk they're willing to take on. SpaceX's bets certainly aren't for the faint of heart".

The company's success will ultimately depend on whether orbital data centers become a reality and whether SpaceX can capture meaningful market share in the AI infrastructure space. Until then, the company's massive AI losses will remain one of the most closely watched metrics in the tech industry.