Here's the question every small satellite operator is quietly panicking about this week: what do you do when the company that built the modern rocket business simply stops selling you a seat? That's our main story today — SpaceX has effectively closed the books on new dedicated Falcon 9 rideshare bookings past 2028, and the scramble it's created is reshaping who gets to fly and who gets left waiting. Before we get there, a few headlines: L3Harris ousted its CEO in an abrupt, same-day boardroom shake-up; the first eight satellites in MDA Space's big Globalstar replenishment fleet just reached orbit; and two direct-to-device rivals, Lynk and Omnispace, merged into a new company called Elveo Mobile to take on SpaceX and AST SpaceMobile. Welcome back to Space Stakes, your daily brief on the business of space. It's Monday, August 17, 2026. Let's get into it.

Let's start with a leadership earthquake nobody saw coming. L3Harris ousted chairman and CEO Christopher Kubasik effective immediately on August 17th, after a board investigation — assisted by independent counsel — found conduct by Kubasik inconsistent with the company's code of conduct. The company was quick to say this had nothing to do with financial reporting, internal controls, customer relationships, or operational performance — just personal conduct, and it says it won't elaborate further. That's about all we actually know, and that's worth flagging as its own caveat: a same-day exit with no severance disclosed is unusual for a company this size. Stepping into the top job is Sam Mehta, who ran the space and communications segments generating more than eighty percent of L3Harris revenue. Before L3Harris, Mehta led Advanced Structures at RTX's Collins Aerospace and spent seventeen years at Sikorsky. He's been the executive overseeing the company's push to expand military satellite and infrared payload production as the Pentagon ramps up spending on the space-based elements of the Golden Dome missile defense initiative — so putting a space and missile-defense executive in the corner office is a pretty direct signal about where L3Harris thinks its growth actually is. Shares fell roughly three percent pre-market on the news. Worth noting too — L3Harris just reorganized from four segments to three back in January, and it's already delayed the IPO of its Missile Solutions business, the old Aerojet Rocketdyne assets, until at least mid-2027. This is a company mid-transformation that just lost the executive who built it.

Now, some actual hardware news amid all that boardroom drama. The first eight of seventeen satellites in MDA Space's Globalstar replenishment program reached orbit Saturday on a Falcon 9, and they're now in commissioning. These are five-hundred-kilogram platforms built by Rocket Lab under a hundred-and-forty-three-million-dollar contract, integrated by MDA in Montréal — and MDA confirms all eight are generating power and operating normally. That might sound like routine satellite plumbing, but it's a milestone for MDA specifically: this is the company's first turn as prime contractor on a commercial low-Earth-orbit constellation, not just a subcontractor building someone else's hardware. Nine more satellites are still finishing integration before the full replenishment fleet is complete, and commissioning has to wrap before Globalstar customers actually see the benefit, so there's real work left. But delivered, functioning hardware in orbit beats another press release with a rendering — and on a day when we're about to talk about how hard it's gotten to launch anything at all, that's worth sitting with for a second.

Over in the direct-to-device world, AST SpaceMobile just picked up FCC permission to test in the 800 megahertz band — thirty days of special temporary authority to test on up to a hundred commercial devices, using spectrum recently transferred from T-Mobile to Grain Management. Low-band spectrum like this travels farther and punches through buildings better than higher frequencies, which matters enormously for a service trying to fill the dead zones your regular carrier can't reach. This follows an agreement in principle that AT&T, T-Mobile, and Verizon announced back in May to pool spectrum for satellite direct-to-device coverage — though as announced, that deal still needed definitive agreements and closing conditions before it's actually final, so treat it as a direction of travel, not a done deal. But let's be precise about what today's news actually is too: a temporary, non-commercial testing authorization, not a spectrum grant. The FCC isn't assigning AST any rights here, and separate reporting notes the bigger prize — waivers for supplemental coverage from space — stays on hold until Grain Management actually names a satellite partner and files the real paperwork. A promising step. Not a green light.

Meanwhile, the direct-to-device field just got a new name to track. Lynk Global and Omnispace completed their merger and rebranded as Elveo Mobile, combining Omnispace's S-band spectrum with Lynk's existing platform to go after SpaceX and AST SpaceMobile. The combined company starts with nine satellites already in orbit — six from Lynk, plus two low-Earth-orbit and one medium-Earth-orbit spacecraft from Omnispace — and says it's planning a three-hundred-twenty-satellite low-Earth-orbit constellation, with commercial service targeted for early 2028. SES, an investor in both predecessor companies, expanded its strategic stake in Elveo the same day. Here's the caveat worth keeping in your back pocket, though: that three-hundred-twenty-satellite constellation and the 2028 date are stated plans, not funded, built hardware — and neither Lynk nor Omnispace had a working commercial direct-to-device service running before this merger. Combining two companies that haven't cracked commercial service yet doesn't automatically produce one that has. It does give them more spectrum and more capital behind the attempt, which is genuinely something.

And finally, Blue Origin is making an interesting bet in the middle of a crisis of its own. The company is building a second launch pad at its Cape Canaveral complex, designated LC-36B, to support a bigger, still-undeveloped New Glenn variant called the nine-by-four — even as it's still repairing the original pad after New Glenn's explosion during a test back in May. When life gives you an exploded rocket, apparently you build a second pad. Blue Origin says it's confident it'll get the repaired LC-36A pad ready to fly the current New Glenn seven-by-two variant again before the end of the year, but it gave no timeline for when LC-36B, or the nine-by-four rocket itself, will actually be ready. The new pad alone is expected to need something like twenty-five hundred tons of steel for a seven-hundred-foot launch tower. That's a serious infrastructure commitment for a rocket variant that hasn't flown yet, layered on top of recovery from one that already blew up — and it's a preview of exactly the kind of capacity math the whole industry is wrestling with right now, which brings us to today's main story.

Our main story today: the launch bottleneck nobody priced in — who inherits the ride-to-orbit business SpaceX built, and can anyone actually fill the gap before Starship's commercially proven? For years, the story of the launch industry has been unambiguous good news — more rockets, more often, at lower prices than anyone thought possible. Over the last three years, an average of two hundred seventy orbital rockets have launched from Earth annually, more than three times what flew a decade ago. Prices are down, cadence is up, and access to space has never been faster. That's the paradox at the center of today's story, because according to Ars Technica's reporting, corroborated by SpaceNews, the exact same industry is simultaneously in what one analyst flatly calls a panic. Here's what's actually happening. SpaceNews reports that at least nine SpaceX partners and customers say the company has effectively stopped accepting new dedicated Falcon 9 rideshare bookings — the Transporter missions where dozens of small satellites split a launch like passengers splitting a cab fare — past late 2028 or early 2029. SpaceX's manifest for the next couple of years is nearly full, and the company appears to be prioritizing its own Starlink constellation and internal payloads over new outside customers. For an industry that built entire business plans around cheap, frequent Falcon 9 rideshares, that's a real problem, and the evidence is showing up everywhere you look. SEOPS, which runs the Waymaker rideshare program, bought a dedicated Falcon 9 for 2028 and sold the entire mission out within three weeks of announcing it — they've since had to add a second dedicated mission just to handle overflow demand. Telesat, the Canadian satellite operator, has eleven Falcon 9 launches booked for its new constellation and says it's now fielding inquiries from other companies asking to share space on those flights; Telesat's answer, per the reporting, is no. Amazon's Leo constellation has had to throttle back satellite production because so few of the dozens of launches it booked years ago are actually ready to fly. And on an earnings call last week, AST SpaceMobile — which we just talked about — told investors that launch availability, not manufacturing or spectrum, is now the pacing item slowing down its whole constellation. Zoom out, and the demand side explains why this is squeezing so hard: the Commercial Space Federation is projecting demand for thousands of satellite launches a year within a decade, and Analysys Mason has forecast that more than thirty-seven thousand satellites will need to reach orbit between 2023 and 2033 alone. So the paradox is stated — now let's dig into why, and who actually wins from it.

So who inherits this business, and is anyone positioned to fill the gap? Caleb Henry, director of research at Quilty Space, put it about as bluntly as you can. What we have, he told Ars Technica, is an industry in panic. And reading between the anecdotes, his implicit conclusion is that this isn't a temporary hiccup, it's a structural mismatch between the demand curve satellite operators are riding and the supply of dedicated launches available to serve it. Now, before we go further, the skeptic's case deserves real airtime, because this isn't quite the crisis the word crisis implies. This is a capacity and scheduling problem, not a failure — SpaceX isn't grounded, it's just full, and full partly because its own Starlink constellation is eating a huge share of its own launch cadence. And per SpaceNews' reporting, some customers in this market actually expect SpaceX to reopen Falcon 9 rideshare slots if Starship's commercial ramp runs behind schedule, which tells you this squeeze might loosen the moment SpaceX decides its own internal demand has been satisfied. There are also fixes already in motion — Starship itself, eventually flying at a scale that could absorb far more payload than Falcon 9 ever could, plus Rocket Lab's Neutron and Firefly's upgraded Alpha Block Two. The problem is every one of those fixes is unproven on schedule, and unproven on schedule has been the default state of this industry for about as long as it's existed. Here's my read on where this actually lands. The paradox Henry's describing isn't really a contradiction — it's what happens when a market grows faster than the industry serving it expected to. Two hundred seventy launches a year sounds enormous until you set it against a forecast of thousands annually within a decade, or thirty-seven thousand satellites needing rides in this one decade alone; against numbers like that, even a tripled launch cadence looks thin. And the real question this raises isn't whether SpaceX is at fault here — it built the golden age this whole ecosystem now depends on — it's whether anyone else can scale fast enough to matter before Starship either proves itself or doesn't. Neutron and the upgraded Alpha are the obvious next candidates, but neither has flown at the cadence or reliability Falcon 9 spent a decade earning, and you don't inherit trust from a manifest, you earn it flight by flight. What I'll be watching is simple: does SpaceX actually reopen Falcon 9 rideshare slots once Starlink's own deployment need eases, or does Starship's commercial ramp arrive fast enough to free up capacity for everyone else anyway? Either path solves the operator's problem. Neither is guaranteed on the timeline anyone actually needs it, and if neither happens on time, you get exactly what SEOPS is already seeing — smaller players buying entire dedicated rockets just to guarantee a seat, which quietly undoes the rideshare economics that made this era of small satellites possible in the first place. Time for the Hype Check. I'm putting this one at a six. The panic is real, and it's backed by named companies — Telesat, Amazon, AST SpaceMobile — turning down favors and rearranging production schedules around a launch bottleneck; that's not manufactured drama. But crisis is doing some of the heavy lifting in the coverage too — this is a company managing scarcity of its own product, not an industry that's stopped launching, and the fixes on the table are real hardware programs, just not yet flying at the cadence this moment demands. So here's the open question I don't have an answer to yet: when SpaceX does eventually reopen its books, will the smaller operators caught in this squeeze still be around to buy the ride, or will they have already signed with someone else by then?

If today's episode had you rethinking how you picture the launch industry, that's kind of the point — subscribe wherever you're listening so you don't miss where this bottleneck goes next, and if you know someone in satellites, send them this one. 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!