NASA just told Blue Origin to find a different rocket

Blue Origin's CEO says December. NASA's chief says maybe 2028.

Someone is wrong. And the answer decides who goes to the Moon.

Last Thursday, Isaacman went on FOX Business and dropped a phrase that should worry every Blue Origin investor. He said NASA is "de-coupling the lander from the launch vehicle and the pad itself." In plain English: we love your Moon lander, but we're done waiting for your rocket.

Let me explain…

Blue Origin's Blue Moon Mark 2 crew lander is a key piece of Artemis. It's supposed to dock with Orion in Earth orbit on Artemis III, set for mid-2027. And it's supposed to land astronauts on the Moon on Artemis IV in 2028. But the lander rides on New Glenn. And New Glenn just blew up.

So Isaacman wants a backup plan. Find a different rocket for the lander. Keep the lander on schedule. Ditch the dependence on LC-36.

But here's the problem. Blue Moon is huge. There aren't many rockets that can carry it. At the CNBC CEO Council Summit on Monday, Isaacman said the options are "limited" due to the lander's mass and size. He named only two vehicles with enough lift: SpaceX and Blue Origin.

In other words… the backup plan for Blue Origin's lander might be SpaceX's rocket.

That's awkward.

And it gets worse. At that same summit, Isaacman himself said the New Glenn pad "may not be restored until 2028." That's a very different number than the "before the end of this year" that Blue Origin CEO Dave Limp promised last week.

Now, I know what you're thinking. "Limp said the fuel tanks survived. He said the rockets are safe. Why 2028?" Fair question. Multiple industry sources told Ars Technica that 12 to 18 months is the more realistic timeline. The root cause of the explosion still hasn't been found. And until it is, nobody builds anything new on that pad.

Meanwhile, SpaceX launched its 66th mission of the year on Saturday night — including two classified Starshield spy sats for the NRO. It was a rare public mention of a program that has launched 13 batches of military sats in two years. SpaceX is running the IPO road show with one hand and launching spy hardware with the other.

So here's where things stand. Blue Origin has great landers, no working rocket, and a wrecked pad. SpaceX has working rockets, the IPO of the century, and a growing defense business. Tomorrow NASA names the Artemis III crew. Thursday SpaceX prices the IPO. Friday it trades.

The gap between first and second place just got wider.

Again.

While the IPO road show was in full swing, SpaceX flew two classified Starshield sats from Vandenberg on Saturday night alongside a batch of Starlinks. It was a rare public acknowledgment of the program. Starshield is SpaceX's military variant of Starlink — hardened, government-only, built with Northrop Grumman sensors on SpaceX buses. The NRO has launched 13 batches in two years, building what it calls a "proliferated architecture" of hundreds of spy sats in low orbit. The total Starlink constellation now stands at 12,191 launched, 10,555 in orbit. SpaceX isn't just a launch company or an internet company. It's becoming America's spy satellite company too.

On Thursday, the American crew aboard the ISS briefly took shelter inside their SpaceX Dragon capsule while Russian cosmonauts addressed new leaks on the station. NASA said the precautions came the same day that Roscosmos prepared "a more extensive repair operation" but gave no details. The ISS has been dealing with leak issues in the Russian segment for months. The station is 26 years old and showing its age. This is exactly why NASA is funding replacements — Voyager's Starlab, Vast's Haven-1, and Axiom's station modules. The clock on the ISS is running down. Every shelter-in-place event makes the case for commercial stations louder.

The SpaceX road show hit peak intensity this weekend. On Saturday, 125 financial analysts from 21 underwriting banks sat down with the company. Today, SpaceX executives and bankers pitch the deal directly to institutional investors. The book closes Thursday. Pricing is set the night of June 11. Trading starts June 12. Demand is strong: Japan upsized its retail slice by 25% on day two. Fidelity opened access to retail buyers. But Morningstar still values the company at $780 billion — 55% below the $1.77 trillion ask. The biggest question on Wall Street this week isn't "will it sell?" It's "will it hold?"

Here's a story almost no one is covering.

SpaceX needs orbital refueling to make Artemis work. Its Starship lander needs about 12 tanker launches to fill up in orbit before it can reach the Moon. The company was targeting a refueling demo this month — two Starships docking and transferring fuel in space.

But Starship is grounded. The FAA mishap probe from Flight 12 hasn't cleared yet. That demo isn't happening in June.

Meanwhile, Rocket Lab quietly set a date. July 17. A small mission called LOXSAT, launching on an Electron rocket with a Photon sat bus. It will test cryogenic liquid oxygen transfer in space — on a tiny scale, but a real one.

NASA is behind the effort. The agency wants to prove that orbital fuel transfer works before it commits to the full Starship refueling chain. And it picked Rocket Lab to do the pathfinder.

Now, this isn't the same thing as two Starships docking. LOXSAT is small. But it would be the first demo of cryogenic fuel handling in orbit by any company. And if Rocket Lab pulls it off before SpaceX even gets Starship back in the air… that's a headline.

In other words, the most important tech milestone for Moon landing prep might come from the small rocket company, not the big one.

Mark July 17.

Saturday night, SpaceX launched two Starshield sats. Most people didn't notice. That's by design.

Starshield is the classified version of Starlink. Same factory in Redmond, Washington. Same basic sat bus. But with military-grade encryption, hardened electronics, and custom sensor packages built by Northrop Grumman.

Think of it this way. Starlink is the phone you buy at the store. Starshield is the phone the CIA carries. Same shape. Very different guts.

The biggest customer is the National Reconnaissance Office — the spy agency that runs America's surveillance sats. In two years, the NRO has launched 13 batches of Starshield sats. The strategy is called "proliferated architecture." Instead of a few billion-dollar sats that take a decade to build, they launch hundreds of cheap ones fast. If you lose one, the rest keep working. If you lose ten, you launch replacements in weeks.

The old model was the opposite. A single NRO sat could cost $5 billion and take 12 years to reach orbit. Starshield compresses that to months.

And here's the investor angle. Starshield revenue doesn't appear as its own line in the S-1. It's buried in the Space segment, which reported $619 million in Q1 sales and a $662 million loss. But the NRO contract is thought to be worth billions over its life. That cash is flowing — just not in a way most investors can see.