Click to see the full-size image
While Washington built a private empire of ten thousand blinking dots and Moscow only started pushing its first commercial birds skyward this spring, another kind of race quietly went hot above our heads. Whoever owns the low orbit owns the next war – the radios, the drones, the reconnaissance, the eyes and the ears of every brigade on the ground. And by the summer of 2026, the map of that race no longer has one player on it. It has six. Some are catching up. Some are just jamming from above and pretending they aren’t there.
The Monopoly That Refuses to Age
By late June 2026, independent astronomer Jonathan McDowell counted 10,705 Starlink satellites in orbit, 10,689 of them working, out of 12,366 launched since 2019 – a scale that alone accounts for roughly two-thirds of everything humanity keeps active in space right now.
What exactly is Starlink, the global low-orbit satellite network? Here is a clear illustration
The FCC (Federal Communications Commission) has already blessed a further 7,500 Gen-2 birds under order DA-26-36 (January 2026), pushing SpaceX’s authorized cap to 15,000 with an eventual ambition of 42,000. SpaceX ran 74 Falcon 9 missions in the first half of 2026, 59 of them dedicated to Starlink – a dedicated mission every three or four days.
And under it all, quietly humming, sits Starshield – the Pentagon-integrated military variant with hardened crypto and priority steering. Starlink over Crimea has already been switched on and off on Elon Musk’s word, a matter of public record, not conspiracy. That single fact is why every other capital on Earth has stopped debating the question and started launching.
Amazon Tries to Grow Up
Jeff Bezos’s Project Kuiper – recently rebranded Amazon Leo – has 331 satellites in orbit as of early June 2026 and another 36 lined up for launch. Impressive in absolute terms. Catastrophic against the FCC’s original deadline: 1,616 birds by July 30, 2026. Amazon isn’t going to hit it. Not even close.
So the company filed for a waiver, and on June 8 the FCC granted it – the half-constellation deadline slips two years to July 30, 2028, and the full 3,232-satellite target holds at July 30, 2029. Kuiper stays in the game, but the message is unmistakable: Amazon’s rocket cadence cannot match SpaceX’s. Nobody’s can.
Behind Kuiper, Bezos’s other bet – Blue Origin’s TeraWave – is still on the drawing board, penciled in for late 2027 with a headline number of 5,408 satellites and laser optics quoted in terabits. B2B, data centers, governments. A promise for now.
The Russian Bird That Refused to Fly
On March 23, 2026, a Soyuz-2.1B lifted off from Plesetsk with the first 16 serial “Rassvet” satellites of Bureau 1440’s constellation. Add the six earlier experimental birds and Russia’s answer to Starlink stood at 22 units in orbit – until, days after the launch, one of them died. Twenty-one now working.
Russia’s “Rassvet” satellites separate from their carrier rocket
Russia’s Plesetsk Cosmodrome
A “Rassvet” satellite in space with its solar panels deployed
The plan on paper is aggressive: 156 birds by end of 2026, 292 by 2027 (the threshold for commercial service), 318 in 2028, roughly 700 by 2030, and 900 by 2035. Backed by about $1.26 billion in state funding against SpaceX’s ten-thousand-strong active fleet. The technology bet, however, is real – inter-satellite laser links, 5G NTN architecture, plasma thrusters. Bureau 1440’s satellites orbit around 800 km, higher than Starlink’s 550, and the laser inter-satellite mesh is the single feature Western analysts have been unable to independently confirm on either Kuiper or China’s GuoWang.
And then, on the same week Rassvet lost its first satellite, President Putin implied on record that the constellation will be used to control heavy strike drones. Nobody in Moscow bothered to walk that back.
“Russia’s low-orbit satellite constellation ‘will be in no way inferior to Starlink, and may even be better” – Putin, June 2026
China Doesn’t Advertise. China Just Launches
Beijing runs two mega-constellations in parallel and doesn’t bother pretending they compete. GuoWang – the state SatNet program, widely reported to serve the PLA – passed 177 satellites in orbit after the June 17 Long March 12 launch from Wenchang. The declared target: 310 by end of 2026, 900 in 2027, 3,600 per year from 2028 onward, up to 13,000 total.
Qianfan – “Thousand Sails,” the Shanghai-led commercial constellation – hit 200 birds on June 5 and 238 by July 5 after the largest polar-orbit launch to date. Beijing wants 324 by the end of July 2026 alone, and roughly 15,000 across the constellation’s life. On June 9, LandSpace’s Zhuque-2E delivered Qianfan DTC-01, the network’s first direct-to-cell test bird. Google that phrase – direct-to-cell – and remember that Starlink was still calling that feature “future” in mid-2024.
China cut per-satellite manufacturing cost by 96% since the program began. This is the same industrial curve that killed Western dominance in shipbuilding, solar, and drones. No reason to expect the orbit to be different.
Europe Discovers Space, Very Slowly
The European Union’s answer – IRIS² – was signed on December 16, 2024, sold as “strategic sovereignty,” and immediately began slipping. Nominal architecture: 290 satellites (264 LEO, 18 MEO plus 8 backups). Nominal service start: 2030. Full operational capability: 2031-2032. First launches don’t happen until 2029, and Brussels has already had to graft in 66 “early delivery” governmental satellites to plug the delay.
Iceland and Norway signed on this June. The French Armed Forces, meanwhile, quietly bought guaranteed OneWeb capacity to bridge the gap until IRIS² materializes. France doesn’t seem to fully believe its own timeline. Neither does anyone else in Brussels.
OneWeb itself – the British-Indian survivor, now folded into Eutelsat – has more than 600 birds and a working global B2B service, orbiting at 1,200 km. Higher latency, thinner throughput, but real revenue. Canada’s Telesat Lightspeed is building for the Arctic and the Canadian military, with roughly $600 million in federal money. Japan skipped the megaconstellation race entirely and went for the harder trick: fusing GEO satellites, LEO birds, and 5G ground towers into a single NTN fabric.
The Jammer That Rides a Molniya
In June 2026 an international academic team dropped a paper that Washington and Brussels are still trying to process. Wideband receivers in Amsterdam and Trondheim recorded the same interference event on February 11 at 1575.42 MHz – the L1 GPS frequency. Time-difference analysis matched the signal, in the researchers’ own words, to “essentially certainty” against one object: Cosmos 2546, a Russian early-warning bird of the EKS (Unified Space System) constellation, on a Molniya orbit.
The interference has been running since 2019 – the year the first EKS satellites launched. It affects only NATO states. It concentrates on Tuesday through Thursday, during Western European working hours. Since June 2020 it hits China’s BeiDou almost identically. The researchers now attribute the effect not to one bird but to the entire EKS constellation acting collectively.
This is the first documented case in human history of a nation jamming a global navigation system from space. Not from a truck-mounted Krasukha in Kaliningrad. From orbit. Every “mysterious” airliner GPS glitch over the Baltics since 2019 now has a return address in Molniya apogee.
Russian satellites become the first in history to create GPS interference from space; no other country has ever pulled this off
The Kill Switch, the Laser, and the Swarm
The strategic question for the next five years is not who has more birds. It is who owns the kill switch.
Starlink is a swarm of ten thousand. Rassvet is a fleet of twenty-one climbing to hundreds. GuoWang and Qianfan together are approaching five hundred and doubling roughly every twelve months. Kuiper is a project. IRIS² is a schedule slip. TeraWave is a PDF.
But the real prize sits in three overlapping questions. Who fields a fully laser-linked constellation first – one that has cut the umbilical cord to vulnerable ground gateways, and can neither be jammed on the RF side nor intercepted on the downlink? Bureau 1440 claims that architecture. SpaceX’s Starshield variants are built for it. Kuiper and GuoWang have not publicly confirmed it. Whoever gets there first buys perhaps three years of orbital dominance before ASAT weapons catch up.
Who keeps sovereign command? Starlink obeys Musk, who obeys Washington, who obeys the moment. Kuiper obeys Bezos. GuoWang, by every Western intelligence assessment, obeys the PLA. Rassvet obeys Russian Bureau 1440, which sits inside sanctioned “IKS Holding.” IRIS² will obey a Brussels committee, which is arguably worse than obeying nobody. Every user in every conflict zone is picking not a broadband provider but a sovereign.
And who first turns the low orbit itself into a weapon? EKS on a Molniya orbit already jams GPS across a continent. Bureau 1440’s next-generation birds are being pitched to steer heavy Russian strike drones. Starshield is already inside the Pentagon’s kill chain. The satellite is no longer a router in the sky. It is a shooter.
Bottom Line
The country without its own LEO constellation in this decade will not just be paying rent to an American billionaire. It will be fighting the next war blind, deaf, and blinking on somebody else’s dashboard. Russia bet late and cheap: twenty-one birds, laser links, a Molniya jammer, and a president saying “maybe better than Starlink” on camera. China bet massive and quiet: four hundred fifteen birds and counting, per-satellite cost cut ninety-six percent, direct-to-cell already flying. Europe bet on a committee.
The orbital chessboard used to have one king. By 2029 it will have three, maybe four. Whoever gets to the fully laser-linked, gateway-free, sovereign swarm first will not need to fire an ASAT weapon to win. They will just quietly stop routing packets for everyone else.
MORE ON THE TOPIC: