The orbital economy just became a three-player game, and two of them are American private companies.
The Summary
- Rocket Lab is acquiring Iridium for $8 billion, positioning itself as the second credible challenger to SpaceX's satellite dominance
- South Korean chipmakers commit $880 billion to chips and data centers, betting national competitiveness runs through AI infrastructure
- Anthropic's Mythos 5 AI model cleared by US regulators after national security review, restoring partial access
The Signal
Rocket Lab's $8 billion Iridium acquisition is the clearest signal yet that the orbital economy is hardening into permanent infrastructure, not science fiction. Iridium brings 75 active satellites and a proven global communications network that already generates revenue. This isn't vaporware. It's cashflow in space.
The deal gives Rocket Lab something SpaceX doesn't prioritize: a mature satellite services business with defense contracts and enterprise customers who need secure, global connectivity today. While Starlink chases consumer broadband and scale, Rocket Lab is buying proven B2B infrastructure and the expertise to run it.
"The orbital economy just became a vertical integration race, and Rocket Lab is betting on owning the full stack."
Here's the strategic tell: Rocket Lab already builds satellites and launches them. Adding Iridium's operational network means they can manufacture, launch, operate, and service their own constellation without touching a third party. That's the playbook. SpaceX does this. Nobody else at scale does this. Until now.
The price tag matters. $8 billion is real money, even in 2026 space economics. It signals two things: institutional investors now believe satellite networks are infrastructure worth utility-like valuations, and Rocket Lab has access to capital markets that treat space companies like Boeing, not like Tesla. This is the maturation of an industry in a single transaction.
Key context:
- Iridium's 75-satellite constellation provides global coverage, including poles
- Defense and aviation sectors rely on Iridium for secure comms
- Rocket Lab's Neutron rocket (designed for mega-constellations) launches in 2025
The South Korea chip commitment is the other half of the story. $880 billion from Samsung, SK Hynix, and others isn't defensive spending. It's recognition that AI compute is now a matter of national security and economic sovereignty. You either make the chips that power AI, or you rent them from someone who does. South Korea is choosing ownership.
This connects directly to the orbital economy. Satellite networks will run AI inference at the edge. Low-latency compute in orbit. Samsung and SK Hynix aren't just making chips for data centers on the ground. They're making chips for constellations that don't exist yet. Rocket Lab and SpaceX will need them.
The Implication
The Rocket Lab-Iridium deal is a forcing function. Amazon's Kuiper and other constellation projects now face a competitor with operational infrastructure, not just plans. Expect acquisition activity. Older satellite networks with revenue (Globalstar, Intelsat assets) become strategic targets for anyone trying to compete without spending a decade on R&D.
For AI builders: orbital compute is closer than you think. If you're designing models that need global, low-latency deployment, start thinking about edge inference in space, not just AWS regions. The infrastructure is coming online faster than the software is ready for it.