The global race for space is no longer about who reaches orbit first. It is about who controls what happens there next. Over the past year, governments and companies have quietly shifted from treating low-Earth orbit as a communications layer to viewing it as strategic infrastructure—part power grid, part data center, part geopolitical leverage. That shift became impossible to ignore when Elon Musk publicly acknowledged that SpaceX has applied to deploy a satellite constellation of a scale never before attempted: up to one million satellites.
The application, filed with the US Federal Communications Commission (FCC) and publicly disclosed in recent weeks, signals a potential inflection point for commercial space, artificial intelligence infrastructure, and orbital governance. While China has announced its own large-scale satellite ambitions—reportedly applying in late December for plans involving around 200,000 satellites during its 15th Five-Year Plan period—SpaceX’s proposal would eclipse every existing or planned constellation combined.
Musk amplified the filing by reposting the news and describing the project as merely achieving a “small goal,” a characteristic understatement that nonetheless underscores the scale of what is being proposed. SpaceX has not yet received regulatory approval, but the filing alone has triggered renewed debate over orbital congestion, space safety, and whether national regulators are equipped to manage projects that operate at planetary scale.
What SpaceX is proposing goes far beyond expanding internet access. According to the FCC submission, the company plans to deploy an entirely new constellation designed to function as an “orbital data center network.” These satellites would be interconnected via laser links and integrated with the existing Starlink system, forming a distributed computing and communications platform in space.
The filing states that the satellites would be capable of high-performance computing, including large-scale artificial intelligence inference and advanced data processing. In effect, SpaceX is proposing to move parts of the global AI infrastructure off the ground and into orbit, creating a space-based, distributed computing layer that could support billions of users and future generations of AI models.
This architectural shift—treating orbit as a computational environment rather than merely a transmission medium—is what gives the proposal its strategic weight. Space, in this vision, becomes an extension of Earth’s digital backbone.
Proponents of orbital data centers argue that space offers structural advantages that terrestrial facilities cannot easily match. Solar energy in orbit is abundant and does not require fuel purchases, with utilization efficiency often cited as several times higher than on Earth. Physical space is effectively unlimited compared with land-constrained data campuses, and thermal management can rely on the vacuum of space rather than water-intensive cooling systems.
Some experts estimate that over a ten-year period, maintaining a conventional ground-based data center can cost roughly $167 million, while a space-based facility of comparable scale could cost as little as $8.2 million. While such comparisons depend heavily on assumptions, they help explain why the idea has moved from science fiction into regulatory filings.
Musk has openly tied this vision to SpaceX’s next-generation launch vehicle, Starship. He has previously said the company aims to reach annual launch capacities on the order of 100 million kilograms, enabling the deployment of massive satellite volumes. Under that scenario, SpaceX could theoretically add around 100 gigawatts of AI computing capacity per year, assuming satellites with roughly 100 kilowatts of power each.
Financing is another critical piece of the puzzle. Industry speculation has intensified around a potential SpaceX public offering, described by some analysts as the largest fundraising event in corporate history. Estimates circulating in financial circles suggest a valuation as high as $1.5 trillion—over 10 trillion yuan—if an IPO were to coincide with Musk’s 55th birthday on June 28 this year. SpaceX has not confirmed any listing plans, but the scale of the satellite application has added fuel to those expectations.
The technical design outlined in the FCC filing places the proposed satellites across multiple narrow orbital shells between roughly 500 and 2,000 kilometers above Earth, with each shell spanning no more than about 50 kilometers in altitude. The structure is intended to optimize coordination and reduce conflicts with other constellations.
That design choice has drawn scrutiny. In the past, SpaceX said it would lower around 4,400 Starlink satellites from 550 kilometers to about 480 kilometers to avoid the 500–600 kilometer band used by many crewed spacecraft. The new application, however, again covers those contested altitudes, raising questions about whether earlier adjustments were temporary measures to reserve orbital capacity for future expansion.
Concerns about safety are not theoretical. With nearly 10,000 SpaceX satellites already in orbit, there have been multiple close-call incidents involving other countries’ satellites and even China’s space station, incidents that prompted formal criticism from Beijing. A constellation scaled to hundreds of thousands—or a million—satellites would dramatically increase the density of objects in low-Earth orbit.
Scientists warn that such congestion could push the orbital environment toward the Kessler Syndrome threshold, where cascading collisions generate debris faster than it can be cleared, rendering entire orbital regions unusable. Once triggered, that process could effectively shut down access to key orbital zones for decades.
For these reasons, the SpaceX filing has placed new pressure on US regulators. While the FCC oversees licensing, critics argue that national agencies cannot unilaterally allocate what is effectively a shared global resource. Chinese officials have already stressed that no country or company should monopolize orbital space without bearing responsibility for long-term safety and sustainability.
It is also important to note that SpaceX’s application remains just that—an application. Approval is not guaranteed, and even with regulatory clearance, the practical constraints are immense. Starship, though technologically ambitious, has yet to demonstrate sustained operational readiness for such missions. Manufacturing satellites at the required pace would itself test the limits of industrial capacity.
Yet the filing has already achieved something significant. It has reframed the conversation about space from exploration and connectivity to control over computational infrastructure beyond Earth. Whether SpaceX ultimately launches tens of thousands of satellites or a fraction of its stated ambition, the direction of travel is clear.
For rival space powers, including China, the message is unambiguous. The competition is no longer just about catching up—it is about preventing the extension of terrestrial technological dominance into orbit. In that sense, the most consequential aspect of SpaceX’s million-satellite proposal may not be whether it succeeds, but how it forces governments and regulators to confront the future of space before it arrives.
