Amazon Bought the Sky. Your SOS Runs Through Bezos Now.

There's a device in your pocket that can talk to a satellite hanging in low Earth orbit and scream your location into the void when you're bleeding out in a ditch somewhere with no cell signal, and that is, genuinely, one of the most astonishing things humans have ever built. A rock in space. Your pocket. A life saved. The engineering that makes that possible should make you stop and feel something.

Now sit with this: the FCC just approved Amazon's acquisition of Globalstar for $11.5 billion, which means the constellation of satellites that Apple's emergency SOS feature runs through is now part of the same company that wants to sell you paper towels, stream you television, and rent you the server that hosts your ex-girlfriend's startup. Your emergency signal, the one that goes up when everything else has gone dark, now travels through Bezos's infrastructure. The feature Apple markets as a safety net runs through his pipes. That's not a hypothetical. That's the deal. That's done.

The orbit is getting crowded in ways that should make you nervous even before you get to the ownership question. Eleven thousand Starlink satellites up there already, four hundred Globalstar Leo satellites, more coming, everyone racing to plant a flag in the sky because the sky is, it turns out, real estate, and real estate goes to whoever can afford the filing fees and the launch costs and the lawyers who know how to talk to the FCC in the right register. The sky was supposed to be the one thing we held in common. It kept not being that. Now it's less that than it has ever been, and the three richest men alive have more say over what happens up there than every government that isn't actively hostile to them combined.

Here's what nobody wants to say out loud because it requires holding two things in your head at once: the technology is not the villain, and the incentive structure absolutely is. Emergency SOS working is genuinely good. A person lost in the backcountry getting found because a satellite heard them is good, full stop, no asterisk. The problem lives one level up, in the question of who gets to decide what that constellation does next, and whether "keep the emergency services running at full capacity even when the revenue doesn't justify it" is a decision a for-profit acquisition is structurally capable of making. The answer, historically, is no. Not because the people running these companies are mustache-twirling monsters (some of them are, but that's not the point), but because the incentive is to deploy spectrum where it makes money, and "someone is dying in a canyon in rural Montana" is not a high-margin use case.

Amazon's Leo project wants that spectrum for commercial broadband. That's the play. Emergency SOS is a feature Apple sells, a cost center for whoever runs the constellation, and a regulatory goodwill chip to wave at the FCC when the deal needs approving. What happens when commercial broadband demand pulls the infrastructure in a different direction? What happens when the constellation gets reoriented around the business case and emergency coverage thins out in the places that need it most (which are, not coincidentally, the places that can't afford to be loud about it politically)? Nobody in the approval process seems to have been in a hurry to answer that, or to put a mechanism in place that would force the answer before the question becomes a tragedy.

The FCC's job, in theory, is to hold the line on exactly this kind of thing. To say: you can have the spectrum, you can do the deal, here are the conditions that make the public interest part of the contract and not just the press release. What actually happened looks more like a rubber stamp with some paperwork attached. The public interest got mentioned. It did not get enforced. It rarely does when eleven and a half billion dollars is on the table and the company writing the check has enough lobbying infrastructure to staff a small country.

What we're watching is the slow, quiet transfer of emergency infrastructure into private hands, one acquisition at a time, with each individual deal looking defensible on its own terms and the aggregate picture only becoming visible if you're paying attention to the pattern instead of the press release. Globalstar wasn't the first. It won't be the last. The sky is filling up with private metal, and the question of what obligations come with that ownership keeps getting answered in the narrowest possible way, in the language of commerce, in the fine print, in the approval documents that nobody reads until something fails.

The technology is remarkable and the deal is a problem and both of those things are true at the same time, and the way we keep letting "the technology is remarkable" close the conversation is part of how we end up here, looking up at a sky we no longer share, hoping the debris misses the things we need.