SpaceX and the AI Infrastructure Frontier: A New Growth Story for the Space Economy

For most of its history, the investment case for SpaceX rested on rockets. Cheaper access to orbit. Reusability. The relentless cadence of Falcon 9 launches that rewired the economics of the launch market. That story remains powerful, but it is no longer the whole story.

What has emerged in 2026 is something more interesting and, for investors thinking about the space economy, potentially more consequential. SpaceX is repositioning itself as a foundational layer of the artificial intelligence infrastructure buildout. The same company that commoditized orbital access is now competing to provide orbital compute, and the same satellite constellation that delivers broadband to fishing boats in the South Pacific could now be the backbone connecting AI workloads between orbit and the ground.

For investors seeking exposure to this convergence, the WisdomTree Space Economy Fund (WSPC) is designed to capture

this evolving story. At launch on July 9, 2026, SpaceX was 14.25% of the fund’s portfolio,¹ making it a cornerstone holding and a lens through which to understand where the space economy is heading.

The AI Demand Signal That Changes Everything

Understanding why SpaceX’s growth trajectory has shifted requires understanding what may be the most important capital allocation trend of this era: AI infrastructure spending.

The hyperscalers, specifically such companies as Microsoft, Google, Amazon, Meta and Oracle, are collectively spending hundreds of billions of dollars annually building out the compute infrastructure required to train and run AI models.² They are running into a fundamental constraint: power. Grid-connected land in the right locations, with permitting and reliable electricity, is increasingly scarce. Every new data center is a multi-year project measured in gigawatts and billions of dollars.

This is where space enters the equation in a new way. SpaceX’s proposed Orbital Data Center system, for which it filed with the Federal Communications Commission (FCC) in January 2026,³ envisions a constellation of up to one million satellites carrying AI compute into low Earth orbit. The logic is elegant, in that solar power in orbit is abundant, continuous, and unconstrained by terrestrial grid limitations. The filing projects that launching one million tons of satellites annually would generate 100 gigawatts of AI compute capacity, a figure equivalent to roughly 20% of current U.S. electricity consumption⁴ dedicated entirely to artificial intelligence.

Figure 1: Artist’s Rendering of a SpaceX AI Satellite

The AI1 satellite, unveiled on June 8, 2026, gave that vision a physical form. With a 70-meter wingspan and 150 kilowatts of power generation, it is designed as a self-contained AI compute node. Musk has framed the technical challenge deliberately:⁵

“Think of it as a rack of compute in space.”

The initial design supports NVIDIA’s GB300 and Vera Rubin platforms, as well as Google tensor processing units (TPUs). The laser inter-satellite link technology powering it is adapted directly from proven Starlink hardware, de-risking the engineering path considerably.

But here is the crucial insight for investors:

Every one of those satellites, whether a Starlink broadband unit or a future AI compute node, needs to reach orbit. SpaceX launches them. The AI infrastructure build in space is, at its core, a launch demand story.

Starlink as the Foundation—and the Funding Engine⁶

Before SpaceX can build orbital data centers, it needs Starlink, which, as of Q1 2026, serves more than 10.3 million active customers across 164 countries, and has become the financial engine that makes everything else at SpaceX possible.

Starlink accounted for approximately 61% of SpaceX’s total company revenue in 2025, generating $11.4 billion, up roughly 50% year-over-year from 2024. Operating margin in the Starlink segment ran at approximately 39%, the only SpaceX division to post consistent generally accepted accounting principles (GAAP) profitability, providing the cash flow foundation that underwrites Starship development and now AI infrastructure investment.

The constellation today numbers more than 10,400 satellites in orbit, of which more than 10,390 are functioning, according to astronomer Jonathan McDowell’s tracking data as of June 1, 2026.⁷ SpaceX has regulatory authorization to deploy more than 40,000 satellites over time. Every expansion phase, the third-generation satellites requiring Starship for deployment, the ongoing reconfiguration of approximately 4,400 satellites from 550 kilometers to 480 kilometers for space safety reasons, the laser inter-satellite links enabling mesh connectivity, requires sustained, high-cadence, reliable launches.⁸

The AI orbital data center ambition stacks on top of this existing launch cadence. According to Musk’s June 8, 2026 presentation and SpaceX’s S-1 filing, SpaceX is targeting an annualized power output of one gigawatt from its orbital AI constellation by end of 2027, scaling to 10 gigawatts and ultimately 100 gigawatts over a longer horizon.⁹ Each of those milestones requires deploying additional satellites at scale. Starship, currently the only launch vehicle capable of the payload economics necessary for this program, is SpaceX’s rocket.

This is vertical integration as a strategic moat. SpaceX controls:

1. The launch vehicle
2. The satellite design and manufacturing
3. The constellation operations
4. The AI workloads running on top of it

The xAI Acquisition: Making the AI Ambition Concrete

In February 2026, SpaceX completed its merger with xAI, the company Elon Musk founded to develop AI models including Grok, at a combined valuation of $1.25 trillion, with SpaceX valued at $1 trillion and xAI at $250 billion. The transaction, described by CNBC as the largest private merger in history, brought xAI’s data centers and the Grok large language model family under the SpaceX corporate structure, along with the social platform X, which xAI had itself acquired in March 2025.¹⁰

For investors, the strategic rationale is straightforward: SpaceX needed AI infrastructure to make its orbital compute vision credible. Similarly, xAI needed a capital base and infrastructure backbone to compete with OpenAI and Anthropic. Together, the combined entity can point to the Colossus supercomputer cluster in Memphis as a working proof point, existing AI revenue, and a roadmap connecting terrestrial compute today to orbital compute tomorrow.

The commercial traction has arrived quickly, and from an unexpected direction. On May 6, 2026, Anthropic, one of the leading AI labs and a direct competitor to xAI’s Grok, announced a deal with SpaceX to use all of the compute capacity at the Colossus 1 data center in Memphis, Tennessee, gaining access to more than 300 megawatts of capacity across more than 220,000 NVIDIA GPUs. The financial terms, subsequently disclosed in SpaceX’s S-1 filing with the SEC on May 20, 2026, are striking:¹¹

• Anthropic will pay $1.25 billion per month through May 2029, with the deal potentially bringing SpaceX over $40 billion in revenue.
• As part of the agreement, Anthropic also expressed interest in partnering with SpaceX to develop multiple gigawatts of orbital AI compute capacity, a detail that connects the terrestrial revenue story directly to the longer-horizon orbital infrastructure thesis.

Then on June 5, 2026, SpaceX disclosed in an amended S-1 filing that Google had signed its own compute agreement, paying $920 million per month from October 2026 through June 2029 for access to approximately 110,000 NVIDIA GPUs. Google described the arrangement as “bridge capacity” to meet stronger-than-expected demand for its Gemini Enterprise platform. Together, the Anthropic and Google deals represent a combined revenue commitment exceeding $2 billion per month, and SpaceX’s S-1 stated it expects to enter into additional similar contracts.¹² What the deals validate, collectively, is that enterprise-scale AI buyers, even those who compete directly with xAI, are willing to pay to access SpaceX’s compute infrastructure. That is a meaningful signal about the company’s positioning as the AI infrastructure buildout matures.

SpaceX’s S-1 prospectus, filed in May 2026 ahead of the company’s June 12 IPO, the largest in history, raising $75 billion at an IPO price of $135 per share and a valuation of approximately $1.77 trillion, disclosed that the company is attributing approximately $26.5 trillion of its stated $28.5 trillion total addressable market to artificial intelligence. That framing is aggressive by any measure. But it communicates something important:¹³

SpaceX’s management views the AI infrastructure opportunity as the primary driver of long-term growth, not an ancillary business.

A Framework for Thinking About SpaceX’s Growth Path

Investors should think about SpaceX’s AI opportunity in three time horizons.

• Near term (2026–2028): Terrestrial AI infrastructure monetization. The Colossus data center and xAI’s model portfolio are revenue-generating today. Enterprise compute contracts like the Google deal provide predictable cash flow. These are conventional AI infrastructure economics, differentiated by Musk’s brand and xAI’s proprietary model development.
• Medium term (2027–2030): Orbital compute proof of concept. The first AI1 satellites reaching orbit and demonstrating commercial workloads in space would be a significant inflection point. Starship’s role in enabling the economics of this deployment phase makes progress on Starship directly relevant to the AI infrastructure investment thesis, not just to launch market share.
• Longer term (2030+): The orbital AI compute scale play. If SpaceX can deploy AI compute capacity in orbit at the scale its FCC filing contemplates, the power constraint problem for AI infrastructure changes in character. This is a decades-long thesis, and the uncertainty is commensurate. But the strategic option value is genuine, and no other company controls all the pieces simultaneously.

For investors, the relevant question is not whether every element of this roadmap will execute on schedule. Very few long-horizon technology infrastructure stories do. The question is whether SpaceX has demonstrated the pattern of capability, which means in manufacturing, in launch cadence, in constellation operations, and in AI model development, that makes the path credible. The track record, from the near-bankruptcy of the Falcon 1 program to becoming the dominant global launch provider, suggests the company has earned meaningful benefit of the doubt on ambitious roadmaps.

Why WSPC?

The WisdomTree Space Economy Fund (WSPC) is designed for investors who believe that space is transitioning from a government-funded research domain to a commercial infrastructure layer, and that this transition is accelerating because of AI.

With SpaceX at greater than 14% of the fund’s portfolio at launch, WSPC offers meaningful, actively structured exposure to the company that sits at the center of this convergence. But the fund’s broader portfolio also captures the ecosystem around SpaceX: satellite manufacturers, launch services providers, space-based connectivity businesses, and defense and civil space contractors who benefit from the rising tide of commercial space activity.

The AI story has changed the frame for the entire sector. The largest constraint on AI scaling is not models or algorithms, but physical infrastructure. Power, compute, and connectivity. Space is increasingly part of the answer to all three. WSPC is positioned to give investors exposure to this thesis in a diversified, transparent, and professionally managed structure.

The space economy has always been a long game. What is new in 2026 is that the demand signal from AI has made it an urgent one.

¹ Source: WSPC fund page, with data as of July 9, 2026. Holdings subject to change.

² Source: Futurum Group. (2026, February 12). AI capex 2026: The $690B infrastructure sprint.

³ Source: Space Exploration Holdings, LLC. (2026, January 30). Application for authority to launch and operate the SpaceX Orbital Data Center System [ICFS File No. SAT-LOA-20260108-00016, Call Sign S00798]. Federal Communications Commission.

⁴ Source: Space Exploration Technologies Corp. (2026, May 20). Form S-1 registration statement under the Securities Act of 1933. U.S. Securities and Exchange Commission.

⁵ Source: Baumgartner, J. (2026, June 9). Musk: ‘Magic’ not needed for SpaceX’s orbital AI data center plan. Light Reading.

⁶ Unless otherwise noted, sourcing for this section is: Space Exploration Technologies Corp. (2026, May 20). Form S-1 registration statement under the Securities Act of 1933. U.S. Securities and Exchange Commission.

⁷ Source: Howell, E. (2026, June 1). Starlink satellites: Facts, tracking and impact on astronomy. Space.com.

⁸ Source: Foust, J. (2026, January 5). SpaceX to lower orbits of some Starlink satellites. SpaceNews.

⁹ Source: Baumgartner, J. (2026, June 9). Musk: ‘Magic’ not needed for SpaceX’s orbital AI data center plan. Light Reading.

¹⁰ Source: Grush, L., Ludlow, E., & Baker, L. (2026, February 2). Elon Musk’s SpaceX combines with xAI at $1.25 trillion valuation. Bloomberg.

¹¹ Sources: xAI. (2026, May 6). New compute partnership with Anthropic; Menn, J., & Somerville, H. (2026, May 20). Anthropic will pay xAI $1.25B per month for compute. TechCrunch.

¹² Source: Roof, K. (2026, June 5). Google will pay SpaceX $920M per month for compute. TechCrunch.

¹³ Source: Kolodny, L., & Levy, A. (2026, June 11). SpaceX raising $75 billion in record-setting IPO as Nasdaq debut awaits. CNBC.