In 2026, astronomer Samantha Lawler and her colleagues placed a number against an ambition — one million orbital AI satellites — and found that the night sky itself hangs in the balance. If SpaceX's proposed mega-constellation were deployed at the scales described, artificial objects could outnumber the stars visible to the naked eye, rewriting the oldest relationship humanity has with the dark above. The study does not adjudicate the future, but it names what is at stake: not merely scientific access to the cosmos, but the unbroken thread of human wonder that has looked upward since the begin
Study: SpaceX's Orbital AI Satellites Could Outnumber Visible Stars
Satellites could outnumber the visible stars in the night sky
Mark
So Lawler's team actually modeled this out—they didn't just say "this seems bad." They ran the math on what one million satellites would look like from Earth?
Mimi
Exactly. They took SpaceX's stated specifications for size and reflectivity and calculated how many would be bright enough to see with the naked eye. The result was that artificial satellites could outnumber natural stars.
Luke
Wait—outnumber the stars that are currently visible to the naked eye, or outnumber all stars? Because there's a big difference in what that claim actually means.
Mimi
The visible stars—the ones you can see without a telescope on a clear night. That's the comparison they made.
Mark
And that matters because astronomers use the night sky to do their work. This isn't just about aesthetics.
Mimi
Right. Professional observation becomes much harder when your field of view is crowded with bright artificial objects. But it's also about something deeper—the night sky as a human experience, something that's been constant for millennia.
Luke
The study shows the problem, but does it say whether SpaceX has actually committed to launching all one million? Or is this a worst-case scenario based on their proposal?
Mimi
It's based on their proposal. Whether they actually deploy that many is still an open question. But the point is that the regulatory frameworks we have now don't seem equipped to handle this scale.
Mark
So this becomes a policy fight, not just a technical one.
Mimi
It has to. You can't solve this with engineering alone. Someone has to decide whether this is allowed.
Luke
And right now, there's no clear answer to that question.
Le Pouls
- A new study calculates that SpaceX's proposed one million orbital AI satellites could make artificial lights more numerous than visible stars — a threshold that would permanently alter the night sky.
- The announcement sharpens a conflict already underway, as Starlink trails have disrupted astronomical observation for years, and this proposal represents an entirely different order of magnitude.
- Beyond light pollution, the sheer scale raises urgent concerns about orbital debris, collision cascades, and the engineering impossibility of tracking a constellation this vast.
- Existing regulatory frameworks — including environmental oversight mechanisms — appear unprepared for a scenario in which a single company could claim and fill this much of near-Earth space.
- The decision about whether megaconstellations of this scale should be permitted is no longer hypothetical; it is moving toward regulators, agencies, and policymakers who may not yet have the tools to answer it.
In 2026, astronomer Samantha Lawler and her colleagues placed a number against an ambition — one million orbital AI satellites — and found that the night sky itself hangs in the balance. If SpaceX's proposed mega-constellation were deployed at the scales described, artificial objects could outnumber the stars visible to the naked eye, rewriting the oldest relationship humanity has with the dark above. The study does not adjudicate the future, but it names what is at stake: not merely scientific access to the cosmos, but the unbroken thread of human wonder that has looked upward since the beginning of recorded time.
In 2026, astronomer Samantha Lawler and two colleagues set out to model something that had begun to feel less like speculation and more like an approaching reality: what would Earth's night sky look like if SpaceX launched one million orbital AI data-center satellites? The answer they arrived at was unambiguous. At the sizes and reflectivity levels proposed, naked-eye satellites could outnumber the stars visible without a telescope — not marginally, but categorically.
The finding sharpened a tension that had been building since Starlink first began populating the sky with moving points of light. But Starlink, vast as it is, represents only a fraction of what a million-satellite constellation would mean. This is not a scaling of the existing problem. It is a different problem entirely.
The implications extend in several directions at once. Light pollution would degrade not only professional astronomy but the simple human experience of darkness — something that has existed, uninterrupted, for all of recorded history. Orbital debris and collision risk compound the concern, as managing a constellation of this size strains the limits of what tracking systems can handle. And the regulatory question looms largest of all: what frameworks govern how much of near-Earth space a single company may claim?
Lawler's study does not offer a resolution. It offers a mirror. If the proposal moves forward as described, this is what the night will become. Whether that future is acceptable — and who gets to decide — remains a question being contested in observatories, regulatory agencies, and boardrooms, with no clear answer yet in sight.
Astronomer Samantha Lawler sat down in 2026 with two colleagues to run the numbers on something that had begun to feel inevitable: what would happen if SpaceX actually launched one million orbital AI data-centre satellites into Earth's orbit. The question was not whether the company had announced such a plan—it had. The question was what it would look like from the ground, from a telescope, from anywhere a person might look up at night.
What they found was stark. If those satellites were deployed at the sizes and reflectivity levels SpaceX had proposed, the researchers calculated that naked-eye satellites—the ones bright enough to see without a telescope—could outnumber the visible stars in the night sky. Not by a small margin. Not as a minor inconvenience. The sheer volume of artificial objects would fundamentally alter what humans see when they look up.
The study crystallized a tension that had been building for years. SpaceX's Starlink constellation had already begun changing the night sky, with astronomers and casual stargazers alike reporting trails of bright points moving across their observations and photographs. But Starlink, as massive as it is, represents only a fraction of what an orbital AI network at the scale Lawler's team modeled would entail. A million satellites is not a refinement of the current problem. It is a different category of problem entirely.
The implications ripple outward in multiple directions. There is the immediate question of light pollution—the degradation of the night sky itself, which affects not only professional astronomy but also the human experience of darkness that has existed for all of recorded history. There is the question of orbital debris and collision risk, the practical engineering challenge of maintaining a constellation so vast that tracking and managing each object becomes exponentially harder. And there is the regulatory question: what frameworks exist, or should exist, to govern how much of near-Earth space a single company can claim and fill?
Lawler's work suggested that the current environmental and regulatory structures—including the National Environmental Policy Act and other oversight mechanisms—may not have been designed with a scenario like this in mind. The question of whether astronomy and megaconstellations can coexist at this scale is no longer theoretical. It is a choice that will have to be made, likely soon, by people in positions to make it.
The study does not propose a solution. It simply holds up a mirror: if SpaceX launches what it has proposed, this is what the night will look like. Whether that future is acceptable, whether it should be permitted, whether there are compromises that could preserve both the company's ambitions and the astronomer's ability to see the cosmos—those are questions still being asked, still being fought over, in boardrooms and regulatory agencies and observatories around the world.
Citations marquantes
The study found that if deployed as proposed, naked-eye satellites could outnumber visible stars— Samantha Lawler's research modeling
Envie de l'histoire complète ?
Lire l'original sur Google News ↗
Politics & Governance