SpaceX is no longer primarily a rocket company. As the aerospace giant prepares for what could be the largest initial public offering in history, a dramatic shift in where the company invests its money and where investors expect future growth reveals a fundamental transformation: artificial intelligence infrastructure, not launch services, is becoming the primary driver of valuation.

The $1.8 trillion valuation reflects investor expectations that SpaceX will raise approximately $75 billion in new capital. But here's what makes this IPO unusual: the company now operates across three distinct business lines, and they're not equally important to the investment thesis anymore. Launch services, which founded the company in 2002, have become supporting infrastructure. Starlink, the satellite internet network, generates the majority of current revenue and profitability. Yet artificial intelligence infrastructure through xAI is increasingly dominating both strategic narrative and capital allocation decisions.

How Does SpaceX's Business Model Actually Work Now?

• Launch Services: Once the core business, rocket launches now primarily support the broader ecosystem rather than serve as the main economic driver themselves.
• Starlink Connectivity: The satellite internet network with more than 10 million subscribers across 164 countries and territories generates the majority of revenue, profitability, and free cash flow that funds other investments.
• xAI Infrastructure: GPU clusters, data centers, and AI compute resources represent the fastest-growing capital allocation priority, even though AI currently contributes a relatively small share of revenue.

Why Is AI Infrastructure Getting More Investment Than the Profitable Businesses?

The contrast is striking. AI-related capital expenditure now materially exceeds combined spending across launch and Starlink infrastructure. This isn't because AI is currently more profitable; it's because management believes the addressable market for AI infrastructure is vastly larger. In the IPO filing, SpaceX estimated an addressable market of approximately $26.5 trillion across AI infrastructure, consumer subscriptions, digital advertising, and enterprise applications, substantially larger than estimated opportunities in launch services and connectivity.

Think of it this way: Starlink is the cash cow that funds the bet on AI. The company is using profits and cash flow from a mature, profitable satellite internet business to build one of the largest AI infrastructure buildouts in the world. This strategy resembles how mature technology platforms have historically funded moonshot ventures, except in this case, the moonshot is artificial intelligence compute capacity.

The scale of this ambition is visible in Colossus, SpaceX's rapidly expanding AI supercomputer. The system reached approximately 200,000 graphics processing units (GPUs) within a remarkably short timeframe, demonstrating an ability to deploy infrastructure at a pace few competitors have matched. For context, a GPU is a specialized chip designed to handle the massive parallel computations required to train and run large language models, the AI systems behind tools like ChatGPT and Grok.

Is xAI Actually Competitive With Other AI Companies?

Here's where the investment thesis gets more complicated. Unlike launch services and satellite communications, where SpaceX possesses clear cost and scale advantages built over decades, the long-term competitive positioning of Grok, SpaceX's AI model, remains less proven. Grok is xAI's large language model, designed to answer questions and generate text, but it faces intense competition from established players like OpenAI's GPT models and Anthropic's Claude.

However, SpaceX appears to be betting less on Grok's dominance and more on the infrastructure layer itself. An important proof point emerged through an agreement with Anthropic, the AI safety company behind Claude. Rather than competing directly with Grok, Anthropic is paying for access to SpaceX's infrastructure, demonstrating that third parties are willing to utilize the company's GPU clusters and compute resources. This arrangement highlights robust demand for AI compute capacity and suggests infrastructure assets may monetize more rapidly than many investors initially expected.

The distinction matters significantly. Building competitive AI models is difficult and uncertain. Building the infrastructure that powers the entire AI ecosystem is a different business with different competitive dynamics. SpaceX increasingly appears focused on the latter, positioning itself as both an AI developer and infrastructure provider.

What Are the Risks Investors Should Know About?

The IPO filing reveals governance structures that give Elon Musk effective control through supervoting shares, limiting minority shareholder influence over strategic decisions. This means investors are backing not only the company's operations but also Musk's long-term capital allocation decisions across launch, communications, and AI infrastructure. The structure introduces governance risks including related-party transactions, concentrated control, and limited shareholder oversight.

At the same time, many investors view Musk as one of the company's key strategic assets given his role in scaling reusable launch technology, Starlink, and broader ecosystem development. Governance is likely to remain both a key attraction and a major risk factor within the broader investment case.

The fundamental challenge is this: much of SpaceX's future valuation opportunity appears tied to successful execution within AI infrastructure and applications. At higher valuation levels, investors are increasingly underwriting the success of xAI and management's ability to commercialize new opportunities beyond the company's existing businesses. That's a bet on both technology execution and capital allocation discipline from a single leader with significant control.

SpaceX's transformation from a launch provider to a diversified infrastructure platform spanning rockets, satellite communications, and artificial intelligence reflects broader trends in how technology companies are evolving. The IPO will test whether investors believe the company can execute at scale across all three domains, or whether the AI infrastructure bet will ultimately define the company's future.