Amazon Bought Falcon 9 Flights From SpaceX. That Is the Story.

Amazon has deployed 302 satellites for its Kuiper broadband constellation, but the milestone that matters landed quietly this week: the FAA issued an indefinite launch moratorium for Blue Origin's New Glenn rocket, grounding the second of Amazon's two primary launch vehicles. Vulcan has been grounded since February. Amazon's only operational contracted rocket is now SpaceX's Falcon 9, the same vehicle that powers Starlink, the service Amazon is spending billions to compete with.

The timing is brutal. The FCC requires Amazon to have 1,618 satellites aloft by July 30. Amazon expects to reach roughly 700. It has filed for a two-year extension or a waiver of the milestone entirely. The company that deliberately excluded SpaceX from its 2022 launch procurement is now buying Falcon 9 flights — not as a backup, but as the backbone of a deployment plan that depends entirely on one rocket, from one company, with 90 days to show it can perform at scale.

Vulcan has been grounded since a February launch in which one of its four solid rocket boosters shed debris. ULA has not announced a return-to-flight date. New Glenn suffered an upper stage failure on its third flight April 19, stranding an AST SpaceMobile satellite in a low, unrecoverable orbit. Blue Origin CEO Dave Limp said one of the two BE-3U engines failed to produce sufficient thrust on a second burn. The company's Second Stage Cleaning and Test facility on the Space Coast also sustained damage in an April incident that preceded the NG-3 failure, damage that reports suggest was not proactively disclosed. Both rockets are essential to Amazon's deployment plan. Neither is flying.

Amazon is now structurally dependent on SpaceX. The Falcon 9 is operational, reliable, and owned by the company Amazon is directly competing with in satellite broadband. Starlink has more than 10,000 active satellites in orbit and serves more than 10 million subscribers in over 100 countries. Kuiper has 302 and is still in limited beta. Amazon has more than 100 launches contracted across providers, but only 10 had occurred as of April. The company says it is targeting 20 or more per year. The math requires every rocket to work. So far, that has not been the case.

The FCC deadline requires Amazon to have 1,618 satellites aloft by July 30. Amazon expects to reach roughly 700, according to filings. It has filed for a two-year extension or a waiver of the milestone entirely. The extension request is a routine regulatory move given the launch constraints. What it reveals is more interesting: Amazon significantly underestimated how hard it is to build a satellite constellation and get it to orbit at scale, on schedule, on a deadline that is now three months away.

The launch bottleneck is not a temporary inconvenience. It is a structural constraint that no amount of contracted capital resolves quickly. Blue Origin is motivated to return New Glenn to flight as fast as the FAA allows — its reputation and its primary commercial customer are both on the line. But fixing a rocket and sustaining a launch cadence are different problems, and Amazon needs the latter more than the former.

What Amazon is actually buying with those Falcon 9 flights is time. Months of orbital deployment that Kuiper cannot afford to lose while SpaceX continues to add subscribers and satellites. The irony is real: the company that wrote a competitive exclusion clause into its launch procurement is writing checks to its primary competitor. That is the story.