$60 Million Won't Break SpaceX's Satellite Grip

Today on Space Stakes: the Space Force just handed five companies sixty million dollars to help them talk to SpaceX's satellite backbone — the real question is whether that buys the Pentagon actual independence, or just insurance. Before that, in the headlines: Ukrainian missiles hit the one factory that builds every Russian Soyuz rocket, SpaceX pulls off a launch record with two Falcon 9s just thirty-eight minutes apart, NASA's moon-landing companies tip off a possible replacement for the aging Lunar Reconnaissance Orbiter, and the Pentagon taps three companies to start designing a way to clean up dead satellites. Welcome back to Space Stakes, your daily brief on the business of space. It's Sunday, August 16, 2026. Let's get into it.

Let's start with SpacePolicyOnline's Marcia Smith, who broke down what's turning into one of the more consequential space stories of the war. Ukraine struck Russia's Progress Rocket and Space Center in Samara, using domestically built Flamingo cruise missiles fired roughly nine hundred kilometers into Russian territory. President Zelensky confirmed the hit himself, calling it a strike on, quote, 'one of the key enterprises within Roscosmos,' end quote. Here's why this matters beyond the war itself: Progress isn't just any factory. It's the only place on Earth that builds the Soyuz-2 family — the rockets that launch Russian crews, cargo, military reconnaissance satellites, and the broadband constellation Moscow's building as its answer to Starlink. There's no backup production line. Russian space expert Anatoly Zak, who runs RussianSpaceWeb.com, says imagery suggests the missiles hit a clean-room building for avionics, right next to vehicle assembly — and he claims, a figure we haven't independently verified, that Soyuz orders were already outpacing what this one plant can build, by as much as thirty to fifty percent for the 2027-to-2030 launch campaign. Now, the damage assessments genuinely conflict — Russian officials call it 'localized,' Ukrainian and open-source channels describe a direct hit and a major fire, and nobody outside has independently confirmed the real production impact yet. Zak's own take: hard to fully destroy production in one strike, but the concern is long-term damage, since Roscosmos has no practical alternative to this system. And remember, Soyuz also flies crew to the International Space Station, where the U.S. and Russian segments remain interdependent.

From a rocket factory under strain to the company setting records with the ones already flying. Saturday night, SpaceX launched two Falcon 9 rockets just thirty-eight minutes apart — one carrying eight Globalstar replenishment satellites out of Cape Canaveral, the other flying the classified USSF-366 mission out of Vandenberg. That beats SpaceX's old turnaround record, set back in August 2024, by twenty-seven minutes. Rocket watchers looking at the booster drop zones think USSF-366 is likely a batch of Starshield satellites — the national-security version of Starlink — though SpaceX, as usual on classified flights, cut its livestream short right after the booster landing. Both boosters landed anyway: one making its fourteenth flight, the other its eighteenth. And the number that actually matters here is ninety-six — that's SpaceX's Falcon 9 launch count for 2026 so far, roughly three-quarters of it Starlink missions. That cadence gap between SpaceX and literally everyone else is basically the subtext running under our whole main story today.

SpaceNews flagged this one too, and it's worth a closer look. Executives from Firefly Aerospace and Intuitive Machines — the two companies that have actually landed spacecraft on the moon under NASA's Commercial Lunar Payload Services program — said on their earnings calls this week that they expect three more CLPS task orders by year's end. Two are landers, business as usual by now. The third, they said, is an orbiter meant to replace the aging Lunar Reconnaissance Orbiter, which has been mapping the moon since 2009 and is currently flying on an extended mission through 2028. Scientists have wanted this for years — a 2022 report called a replacement 'long overdue.' Now, worth flagging clearly: this is industry executives describing an expected award timeline on an earnings call, not a NASA solicitation that's actually been posted, and NASA itself hasn't confirmed a next-gen orbiter is coming. Both companies, notably, already have imaging hardware in the pipeline that could compete for it — Firefly's Ocula service and Intuitive Machines' existing work running LRO's own camera.

And one more before today's main event. SpaceNews reports that the Space Development Agency and the Defense Innovation Unit picked three companies — Firefly, D-Orbit, and Katalyst — for contracts of undisclosed value to study how you'd actually pull a dead, uncooperative satellite out of orbit. According to that same reporting, this is round two of an effort with a track record: a similar 2024 study reportedly led to Starfish Space landing a fifty-two-and-a-half-million-dollar contract to deorbit up to seven spacecraft. We haven't independently confirmed either the new awards or that earlier Starfish contract ourselves, but the pattern of small design studies scaling into real money is at least plausible, going by SpaceNews's account. The honest caveat either way: this is a preliminary design contract, not a funded deorbit mission. Firefly's proposed spacecraft, Elytra, hasn't flown its first mission yet. Katalyst's servicing vehicle is already up there wrestling with an aging satellite, and it actually had attitude-control trouble during setup last month before engineers fixed it with new flight software. This is design money for now — the money that actually counts arrives when SDA and DIU pick one of these three for the real on-orbit demonstration.

Our main story today: sixty million dollars in seed money, five companies, and whether that's real leverage against SpaceX's hold on military satellite communications, or just a symbolic hedge. Here's what happened. Space Systems Command announced August 13th that it's handing contracts to five companies — Amazon Leo, Lockheed Martin, Northrop Grumman, Rocket Lab, and York Space. Each one gets twelve million dollars, split into two pieces: a ten-million-dollar fixed-price prototyping contract, plus a two-million-dollar Other Transaction Authority agreement to build what's called a Space Exchange Point — essentially a router satellite whose job is plugging outside networks into a much bigger backbone. And that bigger backbone is the real story underneath the story. Back in May, SpaceX won a $2.29 billion contract to build the actual Space Data Network — an operational, Starshield-based prototype due by the end of 2027. Starshield, remember, is the military-grade cousin of Starlink. So put the two numbers side by side: two point two nine billion dollars for the core network, versus sixty million split five ways to make sure other companies' hardware can talk to it. That's not close — the interoperability money here is worth something like a fortieth of the backbone contract it's plugging into. Now, one company had a genuinely loaded week inside this story: Rocket Lab. Separately from this sixty-million-dollar award, Rocket Lab landed a $397 million Space Force contract to build and launch what are called 'Flatellite' threat-tracking satellites under a program named SB-AMTI — and those satellites are set to fly on Neutron, Rocket Lab's next rocket, which, worth saying plainly, has not flown yet. On top of that, Rocket Lab also has a previously reported $266 million hypersonic test contract already in its pocket. So in a little over a week, Rocket Lab picked up pieces of both the sensing side and the networking side of this emerging military space architecture — real money, even if some of it is riding on hardware still waiting for its first flight. So why is the Space Force doing any of this at all? The pitch, in plain English, is resilience — the Pentagon doesn't want its entire military satellite communications backbone dependent on one company's proprietary system, especially with everyone thinking hard right now about what happens to space infrastructure during a conflict. 'Multi-vendor' is the term of art here — get enough companies talking to the same network that no single vendor becomes a single point of failure. And the five companies picked span a real cross-section of the industry: a hyperscaler's satellite arm in Amazon Leo, two traditional primes in Lockheed and Northrop, and two newer space companies in Rocket Lab and York.

So here's the deeper read: what does twelve million dollars per company actually buy, when the network they're plugging into was already built and owned by somebody else? Defense Daily's @calvinb21 reported it straight on X, framing it as, quote, 'Space Force Selects Five Companies To Help Leverage Commercial Satellites As Part Of SDN,' end quote — and that word, 'leverage,' is doing a lot of work. These five vendors get paid to plug into the backbone, not to replace it. Not everyone thinks plugging in means real interoperability. Laurie Scott, who posts as @Laurieneuco on X, put it sharply: twelve million dollars per vendor, she wrote, buys interface validation, not flight hardware — and she says she's skeptical the crosslinks connecting these networks are genuinely open, warning that proprietary ports could make the whole mesh a chokepoint no matter how many vendors get a badge. And that's the actual tension here. This is Other Transaction Authority money — a contracting tool the Space Force leans on because it moves faster than a normal procurement, but it's also a mechanism the Government Accountability Office has repeatedly flagged for weaker public reporting. So who's actually checking the math here? Five companies are being paid to prove they can talk to a network SpaceX alone designed, under a contracting mechanism that keeps a lot of that math out of public view. What would actually make this credible? Not another contract announcement — an actual demonstration where a Lockheed or a Rocket Lab satellite routes real data through one of these Space Exchange Points, on the SpaceX-built backbone, on terms those companies didn't have to negotiate satellite by satellite. That's the test. Nobody's published that it's happened, and given these are ten-million-dollar prototyping deals just getting underway, it's nowhere close yet. My own take: twelve million dollars each buys these five companies a real technical foothold and a working relationship with Space Systems Command — genuine value, especially for a company like York or Rocket Lab angling for a bigger piece of military satcom down the road. But sixty million dollars split five ways next to a $2.29 billion prime contract doesn't rebalance power. It's a hedge the Space Force can cite whenever someone in Congress asks whether the military's gotten too dependent on one vendor for its communications. Whether that hedge turns into genuine competition depends entirely on whether these Space Exchange Points end up carrying real traffic, or just sit there as a proof-of-concept line in next year's budget slides. Time for the Hype Check. I'm putting this one at a four out of ten on substance. The contract structure is real, the money's obligated, and Rocket Lab's parallel wins this week show the company's genuinely building a footprint across sensing and networking both. But 'multi-vendor architecture' is carrying a lot of weight for a program that's still sitting at the interface-validation stage, dwarfed nearly forty to one by the backbone contract it's meant to check, and running through a contracting tool that limits what the public actually gets to see. It's real money, but genuine competition here is still theoretical, not demonstrated.

Here's what to watch: the first real demonstration where a Lockheed or Rocket Lab satellite actually routes traffic through the Space Force's backbone — that's the moment this stops being a press release and becomes a fact. And Rocket Lab's Neutron rocket has its own reckoning coming, since a few hundred million dollars of new contract backlog rides on a vehicle that hasn't left the pad yet. If you want the space business explained without the press-release gloss, follow Space Stakes wherever you're listening. This has been Space Stakes, an AI-voiced podcast, created and built by a real human using today's cutting-edge technology. Nothing you heard on this show is financial advice. I'm Brian Lampert, and I'll catch you all tomorrow — take care!