SpaceX wants to put up to one million data centers in orbit. The astronomical community is furious. And the Federal Communications Commission has, so far, done almost nothing to stop it.
A newly surfaced FCC filing from SpaceX proposes a constellation of satellites that would function as orbital computing infrastructure — essentially, server farms in space. The application, which envisions as many as one million units, has triggered alarm among professional astronomers who already struggle with the light pollution created by SpaceX’s existing Starlink broadband constellation. As reported by Futurism, the sheer scale of the proposal dwarfs anything previously contemplated in low Earth orbit, and the potential consequences for ground-based observation could be severe.
The numbers alone are staggering. SpaceX currently operates roughly 7,000 Starlink satellites, already the largest constellation ever deployed. Those satellites have become a persistent headache for observatories worldwide, streaking across long-exposure images and contaminating data sets that took months or years to assemble. Now multiply that problem by more than a hundredfold.
That’s what astronomers are staring down.
The concept of orbital data centers isn’t entirely new. Several startups have floated the idea of moving computation off-planet, citing advantages in cooling — the vacuum of space is an efficient heat sink — and access to uninterrupted solar power. But no company has proposed anything close to this magnitude. SpaceX’s filing with the FCC suggests the company sees orbital computing as a natural extension of its Starlink infrastructure, potentially serving artificial intelligence workloads, cloud computing customers, and government clients who want data processing closer to satellite-based sensors.
The filing itself is thin on technical specifics regarding optical brightness, orbital altitude distribution, and end-of-life deorbiting plans for a million units. That vagueness is part of what has researchers worried. The American Astronomical Society and the International Astronomical Union have spent years negotiating with SpaceX over brightness mitigation for Starlink satellites, achieving modest gains through sunshade visors and darker coatings. Those negotiations were painstaking. They were also voluntary — no U.S. law compels satellite operators to limit the reflectivity of their spacecraft.
And therein lies the core tension. The regulatory framework governing satellite constellations was designed for an era when a few hundred objects in orbit constituted a crowded sky. The FCC’s mandate covers spectrum allocation and interference with radio communications, not optical pollution. The Federal Aviation Administration oversees launch safety. NASA tracks orbital debris. But no single federal agency has clear authority to regulate the brightness of satellites or to weigh the cumulative impact of a million new objects on scientific observation. The gap is enormous, and SpaceX is driving straight through it.
Astronomers have been sounding this alarm for years. Dr. Meredith Rawls, a research scientist at the University of Washington’s Rubin Observatory, has been among the most vocal critics of unchecked constellation growth. The Vera C. Rubin Observatory, currently under construction in Chile, will conduct the Legacy Survey of Space and Time — a decade-long project to map the entire visible sky repeatedly. Its wide-field camera is exquisitely sensitive, which makes it exquisitely vulnerable to satellite streaks. Even with software designed to mask contaminated pixels, the data loss from tens of thousands of satellites is nontrivial. A million orbital data centers would make portions of the survey effectively unusable during certain hours.
SpaceX has pushed back on such concerns in the past, arguing that its engineering teams work cooperatively with the astronomy community and that technological solutions — dimmer satellites, predictive avoidance software, post-processing algorithms — can mitigate the worst effects. The company points to its VisorSat design and later DarkSat experiments as evidence of good faith. But critics note that these measures were implemented only after sustained public pressure, and that SpaceX has never agreed to binding brightness limits.
The stakes extend well beyond aesthetics. Ground-based telescopes remain indispensable for planetary defense — the detection and tracking of near-Earth asteroids that could pose collision risks. The more satellites cluttering the field of view, the harder it becomes to pick out a faint, fast-moving rock against a background of artificial streaks. A 2023 report from the National Academies of Sciences, Engineering, and Medicine warned that satellite mega-constellations could meaningfully degrade planetary defense capabilities if growth continued unchecked.
There’s also the debris question. One million objects in low Earth orbit would dramatically increase collision risk, even if each individual unit is small. The Kessler syndrome — a theoretical cascade in which one collision generates debris that triggers further collisions — becomes less theoretical with every order-of-magnitude increase in orbital population. SpaceX has touted its satellites’ ability to autonomously maneuver to avoid collisions, but autonomous avoidance systems are only as good as the tracking data they rely on, and the U.S. Space Force’s catalog of tracked objects already struggles to keep pace with current traffic.
So who decides whether this happens?
Right now, the answer is largely SpaceX itself, subject to FCC licensing. The commission has approved Starlink in phases, attaching conditions related to spectrum use and orbital debris mitigation but largely deferring on questions of optical impact. Congress has shown little appetite for comprehensive space sustainability legislation, though a handful of bills have been introduced in recent sessions. The Orbital Sustainability Act, proposed in 2024, would have directed the Department of Commerce to develop guidelines for satellite brightness, but it stalled in committee.
Internationally, the picture is no clearer. The United Nations Committee on the Peaceful Uses of Outer Space has discussed the issue but lacks enforcement power. The International Telecommunication Union governs radio spectrum, not reflected sunlight. And the Outer Space Treaty of 1967, the foundational document of space law, was written when the idea of a million commercial satellites would have sounded like science fiction.
Industry observers note that SpaceX’s filing may be partly strategic — staking out spectrum rights and orbital slots that the company may never fully use, a common tactic in satellite licensing. But even a fraction of one million units would represent an unprecedented transformation of the orbital environment. And SpaceX has a track record of building what it files for. When Elon Musk first proposed a constellation of thousands of internet satellites, skeptics dismissed the timeline as fantasy. Starlink now serves millions of customers in dozens of countries.
The AI angle adds commercial urgency. Data center capacity on the ground is constrained by power availability, cooling costs, and permitting battles with local governments. Companies like Microsoft, Amazon, and Google are spending tens of billions of dollars annually on terrestrial data center construction, and demand — driven largely by the computational requirements of large language models and other AI systems — is outstripping supply. If SpaceX can demonstrate that orbital computing is economically viable, the market pull could be immense. The company’s vertically integrated model — it builds the rockets, the satellites, and the ground stations — gives it a cost structure that no competitor can currently match.
But viable for whom? That’s the question astronomers keep asking. The night sky is a shared resource, one that predates any corporation or government. Indigenous communities, amateur astronomers, and billions of people in the developing world have cultural and practical relationships with an unobstructed sky that no cost-benefit analysis fully captures. Professional astronomy, meanwhile, is a public good funded largely by taxpayers. The telescopes threatened by satellite constellations were built with public money to answer fundamental questions about the universe. Allowing a private company to degrade those instruments without meaningful regulatory review represents a transfer of value from the public to a single corporation’s shareholders.
SpaceX would likely counter that its services — global internet access, disaster communications, support for remote and underserved communities — also constitute public goods. That argument has merit. Starlink has provided connectivity in war zones, after natural disasters, and in rural areas where terrestrial broadband is unavailable or prohibitively expensive. The tension between these competing goods is real, and pretending otherwise doesn’t help.
What would help is a regulatory framework that actually accounts for the full range of impacts. The FCC could condition future constellation licenses on binding brightness standards developed in consultation with the astronomical community. Congress could grant the Department of Commerce or another agency explicit authority over the optical and environmental effects of satellite constellations. International bodies could establish norms — even non-binding ones — that create reputational costs for operators who refuse to cooperate.
None of this is happening fast enough.
Meanwhile, SpaceX continues to launch. The company’s Falcon 9 rocket flies roughly every three days, frequently carrying batches of Starlink satellites. Its next-generation Starship vehicle, currently in testing, is designed to carry far larger payloads — potentially hundreds of satellites per flight. If Starship achieves the rapid reusability SpaceX is targeting, the economics of deploying a million orbital units shift from implausible to merely expensive.
The astronomy community isn’t powerless, but its leverage is limited. Researchers can file comments in FCC proceedings, publish studies documenting the impact, and lobby Congress. They can work with SpaceX engineers on mitigation technologies. They can build space-based telescopes that operate above the interference — though those instruments cost billions and take decades to develop. What they cannot do is unilaterally halt the commercialization of low Earth orbit.
And that commercialization is accelerating. Amazon’s Project Kuiper is preparing its own constellation of over 3,000 broadband satellites. OneWeb, now merged with Eutelsat, operates several hundred. China has announced plans for a national mega-constellation called Guowang, potentially comprising 13,000 satellites. Each new entrant compounds the problem. Even if SpaceX were to adopt the most aggressive brightness mitigation available, the cumulative effect of multiple constellations from multiple nations could still overwhelm ground-based astronomy.
The situation has drawn comparisons to earlier environmental conflicts — the unregulated dumping of industrial waste into rivers before the Clean Water Act, or the unchecked emission of chlorofluorocarbons before the Montreal Protocol. In each case, the commons was degraded incrementally by actors pursuing rational self-interest, and the damage became apparent only after significant harm had already occurred. The difference with orbital pollution is speed. Rivers can be cleaned. The ozone layer is recovering. But astronomical data lost to satellite contamination is gone forever, and the observations that data would have enabled — the detection of a potentially hazardous asteroid, the characterization of an exoplanet atmosphere, the measurement of dark energy’s acceleration — cannot be retroactively recovered.
SpaceX’s million-unit filing is, in one sense, just a piece of paper at the FCC. But it signals intent. It tells the market, competitors, and regulators where the company is headed. And it forces a question that American policymakers have been content to defer: Who owns the night sky, and what are we willing to sacrifice to connect it to the cloud?
The astronomers already know their answer. They’re waiting for everyone else to catch up.
WebProNews is an iEntry Publication