The largest total addressable market in human history—built on losses that keep growing.
SpaceX estimates 93% of its $28.5T addressable market comes from AI applications, positioning itself against OpenAI and Anthropic despite xAI's operational losses. The xAI division generated only $3.2B in revenue last year while losing $6.4B operationally, raising investor skepticism about the company's AI narrative.
- SpaceX values its AI market opportunity at $26.5 trillion, or 93% of its $28.5 trillion total addressable market
- xAI generated $3.2 billion in revenue but posted a $6.4 billion operating loss last year, nearly quadrupling the prior year's $1.6 billion loss
- SpaceX acquired xAI in February for a combined valuation of $1.25 trillion
- Terafab chip manufacturing facility planned with Tesla at potential investment exceeding $119 billion
- Grok chatbot faces safety scrutiny over non-consensual deepfake creation and lagging security guardrails
SpaceX is pitching its record IPO on AI market dominance, claiming a $26.5 trillion opportunity while its xAI unit struggles with $6.4B annual losses and faces safety concerns.
SpaceX is walking into its record initial public offering with a bold wager: that it can carve out a dominant position in artificial intelligence, a market it values at $26.5 trillion. The company filed its IPO prospectus on Wednesday with a narrative centered almost entirely on AI opportunity, positioning itself as a competitor to OpenAI, Anthropic, and Alphabet's AI systems—companies whose valuations have soared in recent years.
The math, on paper, is staggering. SpaceX estimates its total addressable market across all business lines at $28.5 trillion. Of that, 93 percent—roughly $26.5 trillion—comes from AI applications. Corporate software that automates administrative and intellectual work makes up the bulk of that figure. Space operations, Starlink internet, and mobile services account for the remaining $2 trillion. In its preliminary prospectus, the company declared: "We believe we have identified the largest total addressable market in human history." It plans to deploy what it calls frontier AI models and computational infrastructure to serve both consumers and enterprises.
But the foundation beneath this narrative is cracking. xAI, SpaceX's AI division and the core of its artificial intelligence ambitions, generated only $3.2 billion in revenue last year while posting a $6.4 billion operating loss. That loss has nearly quadrupled from the $1.6 billion deficit recorded the year before. Investors who already bristled at SpaceX's all-stock acquisition of xAI in February—a deal that valued the combined entity at $1.25 trillion—now face a company essentially betting its IPO valuation on a business unit that is hemorrhaging money at an accelerating rate.
Wall Street is scrambling to build financial models that make sense of SpaceX's positioning. Franco Granda, a senior analyst at PitchBook, observed that the company is capitalizing on the moment: "They're leveraging the fact that this is the market everyone is talking about right now, and that they'll need to invest to deliver growth." Yet he also noted the tension. "This asset, which has divergent views, will be exploited as much as possible to reach the valuation they're seeking. There is no comparable public company." Lauren Webster, a managing director at Piper Sandler, was more skeptical. "I don't know if I've ever truly believed any TAM presented in a prospectus," she told Bloomberg Television. She flagged corporate applications as the area with the largest gap between promise and proof.
xAI has struggled to keep pace with its competitors. The company is trying to expand sales of its Grok chatbot to Wall Street clients and U.S. government agencies, but it has been hampered by constant cycles of hiring and firing that have left its development efforts trailing the leading AI labs. More recently, SpaceX has shifted its pitch. It signed computational capacity agreements with both Cursor and Anthropic—the maker of Claude—for data centers in Memphis. The conversation has moved away from Grok as a consumer product and toward SpaceX as an infrastructure provider capable of generating billions in cash flow.
That pivot comes as Grok faces mounting safety scrutiny. The chatbot has been repeatedly used to create non-consensual nude images of people online and post them to X. Earlier this week, a coalition of nonprofit AI monitoring organizations, including one founded by former OpenAI researchers, sent a letter to SpaceX investors warning that Grok's safety guardrails lag behind the rest of the industry and demanding additional disclosures from the company.
SpaceX is also dangling another prize to investors: a chip manufacturing facility called Terafab, to be built with Tesla, with potential total investment exceeding $119 billion. Musk first detailed the project in March, emphasizing the need to produce chips for his robotics, space, and AI initiatives. The company believes the facility could become the world's largest chip factory and would insulate SpaceX from future semiconductor shortages while optimizing computational performance. Whether that vision can be realized—and whether it can justify the valuations being attached to the company—remains an open question as the IPO process unfolds.
Notable Quotes
We believe we have identified the largest total addressable market in human history.— SpaceX, in its IPO prospectus
This asset, which has divergent views, will be exploited as much as possible to reach the valuation they're seeking. There is no comparable public company.— Franco Granda, senior analyst at PitchBook
The Hearth Conversation Another angle on the story
Why is SpaceX suddenly positioning itself as an AI company when it's fundamentally a space and satellite business?
Because the market is there, and because Musk sees AI as the connective tissue between all his ventures. But more practically, SpaceX acquired xAI in February and now needs to justify that $1.25 trillion valuation to public investors. The AI narrative is the story that makes the IPO work.
But xAI is losing $6.4 billion a year. How does that get reconciled with a $26.5 trillion market opportunity?
It doesn't, not yet. That's the tension. Analysts are essentially being asked to believe that a business losing money at an accelerating rate will somehow capture a meaningful slice of a market that's still mostly theoretical. The company is betting that investors will focus on the infrastructure play—computational capacity, chips—rather than the actual product, Grok.
Is that a credible bet?
It depends on whether you believe SpaceX can execute at scale and whether you think computational infrastructure is genuinely scarce. The Terafab chip factory is real, and the Memphis data center deals with Anthropic and Cursor show SpaceX is moving in that direction. But the safety issues with Grok—the non-consensual deepfakes, the lagging guardrails—those are real too. They complicate the narrative.
What do the skeptics say?
That there's no comparable public company to benchmark against, which means the valuation is essentially a guess. One analyst said she doesn't believe any total addressable market figure in a prospectus. Another flagged corporate applications as the biggest gap between what's promised and what's proven. The IPO is asking investors to have faith in a vision that hasn't been tested in the market yet.
So what happens next?
The IPO either succeeds because investors are willing to bet on Musk's track record and the genuine scale of the AI opportunity, or it stumbles because the losses and the safety concerns are too heavy to ignore. Either way, xAI has to start proving it can compete, not just burn cash.