SpaceX has filed a proposal with the United States Federal Communications Commission (FCC) to launch one million new satellites to serve as orbital data centers for AI computing power, a development that researchers warn will outnumber visible stars and permanently alter the night sky for everyone on Earth.
The push for orbital AI infrastructure represents a massive escalation in megaconstellations. According to reports from The Independent, more than 10,000 Starlink satellites currently orbit the planet, disrupting research telescopes and remaining visible even in remote locations. Now, the scale is shifting from thousands to millions.
Simulating the Orbital Crowding of a Million Satellites
A team of researchers previously published a predictive model when four planned megaconstellations—SpaceX’s Starlink, Amazon’s Kuiper (now Leo), the U.K.’s OneWeb, and China’s Guowang—accounted for 65,000 proposed satellites. That model calibrated observations against real Starlink hardware to predict that one in fifteen visible points in the night sky would be a satellite rather than a star.
A million satellites changes that equation entirely. Because the human eye can see fewer than 4,500 stars in an unpolluted night sky, a constellation of this magnitude means artificial objects will outnumber stars for large portions of the night and the year worldwide. To model this, researchers built an updated simulation leveraging data from astrophysicist Jonathan McDowell’s website to map orbits consistent with the sparse details provided in the SpaceX FCC filing.
The simulation scales brightness models by factoring in size jumps from Starlink V1 and V2 to predicted V3 specifications, alongside higher power and complexity requirements. The resulting projections show that higher orbits will keep these satellites visible for longer periods of the night.
The Greenwashing Debate Surrounding Orbital Data Centers
Terrestrial data centers face intense criticism for consuming massive amounts of electricity and water. SpaceX has pitched moving data processing into space as an environmentally friendlier alternative. However, analysts point out that this framing ignores the heavy environmental costs associated with rocket launches, ongoing orbital operations, and eventual atmospheric re-entry.
Furthermore, the FCC filing lacks crucial operational details. The proposal fails to account for collision risks, atmospheric pollution, or the engineering required to disperse waste heat generated by orbital data centers.
What Lies Ahead for Global Astronomy
The regulatory path for these million satellites runs through the U.S. FCC and the International Telecommunication Union (ITU), where numerous national governments have already filed proposals for large satellite systems. As aerospace firms chase orbital computing dominance, the astronomical community faces a challenge in preserving dark skies against a backdrop of corporate infrastructure.
