The Orbital Power Map: SpaceX, Amazon, and the Race to Own the Next Stack
On March 11, FCC Chairman Brendan Carr posted a public broadside on X directed at Amazon: the company, he wrote, should focus on the fact that it will fall roughly 1,000 satellites short of its own upcoming deployment milestone, rather than spending time and resources filing petitions against companies that are putting thousands of satellites into orbit. He told Reuters and CNBC separately that he does not expect Amazon's petition to gain traction at the Commission. Regulators do not signal this bluntly unless they want the market to understand who currently has process momentum. That is the real headline — everything else is context.
What Amazon Did, and Why It Backfired
On March 6, Amazon's satellite division — Amazon Leo, formerly Project Kuiper — filed a 17-page petition urging the FCC to deny SpaceX's application to launch up to one million solar-powered, AI-focused orbital data center satellites at altitudes of 500–2,000 km. SpaceX had filed the application on January 31, framing it as "the first step toward becoming a Kardashev II-level civilization." The FCC opened a public comment period that drew over 1,200 responses by the March 6 deadline.
Amazon's technical arguments were substantive: the proposal is "speculative, conceptual, and incomplete," lacks basic RF and orbital design details, and — given that only 4,526 satellites were launched globally in all of 2025 — would take centuries to fully deploy. Approving it, Amazon argued, would force every competing LEO operator to plan around a constellation that may never exist, effectively warehousing orbital slots.
The Glass Houses Trap
The sting of Carr's rebuke is structural, not rhetorical. On January 30 — one day before SpaceX even filed its application — Amazon Leo asked the FCC for a 24-month extension to push its July 2026 milestone (requiring ~1,618 satellites in orbit) back to July 2028. Amazon's own filing confessed the cause: a 2023 prototype success triggered unexpected re-engineering that delayed mass manufacturing by nine months; in 2025, only 7 of 20+ planned launches were completed. Most damaging, Amazon disclosed it had purchased 10 additional launches on SpaceX's Falcon 9 — its chief rival's rocket — after delays with Blue Origin and ULA. The company has invested over $10 billion and placed just ~200 satellites in orbit since April 2025.
A company simultaneously seeking regulatory leniency for itself while demanding strictness toward a competitor has surrendered its moral high ground before a Commission already predisposed against it.
The Strategic Option, Not the Business Plan
Sophisticated investors should resist reading the million-satellite filing literally. SpaceX does not need the plan to be physically realistic for it to be strategically successful. The filing accomplishes four things at once: it plants a claim on future spectrum and orbital positioning; it forces rivals into expensive defensive filings; it aligns SpaceX's story with AI capital markets; and it repositions Starlink from a connectivity asset into national compute infrastructure. With reported merger talks between SpaceX and Elon Musk's xAI ahead of a potential SpaceX IPO in 2026 — and xAI having just disclosed a Pentagon AI contract valued up to $200 million — that repositioning carries serious valuation weight.
The real investment variable is Starship cadence. If Starship achieves reliable, high-frequency, low-cost heavy lift, today's absurd space-economics assumptions stop being absurd. If it doesn't, this filing is a negotiating instrument, not a deployment roadmap.
The Hidden Loser and the Real Scarcity
The strategic threat to Amazon Web Services is not orbital compute replacing terrestrial data centers this decade — it won't. The threat is that a vertically integrated SpaceX/xAI stack offering launch, bandwidth, edge connectivity, and specialized compute could permanently carve premium defense and sovereign workloads away from cloud incumbents.
The overlooked winner is launch capacity itself. Amazon's filing inadvertently confirmed the market's suspicion: rocket availability is the genuine bottleneck. Any company with credible exposure to launch infrastructure or satellite manufacturing sits closer to the value chain than any abstract "space platform" narrative.
The biggest underpriced bear case for all LEO operators is not competition. It is orbital governance failure — the point at which constellation density triggers debris cascades that cap the cash-flow duration of every player in the sector, a concern raised by astronomers and scientists among the 1,200+ public comments the FCC received.
The regulatory win and the economic win are not the same thing. SpaceX is winning the narrative, the regulator, and the option value. The million-satellite filing is the opening bid in a decades-long negotiation over who owns the stack above the atmosphere.
not investment advice
Sources: https://x.com/BrendanCarrFCC/status/2031746827645562940