Amazon announced a deal to acquire satellite operator Globalstar, a move that instantly vaults the company into the direct-to-device connectivity market and sets up a collision course with SpaceX’s Starlink. It’s a classic Amazon play: buy what you can’t build fast enough. But the company it’s trying to catch has spent five years proving that in the satellite business, execution beats capital — and SpaceX has both.
The acquisition gives Amazon globally licensed mobile satellite services spectrum, an operational satellite fleet, and ground infrastructure already serving Apple’s emergency messaging features on iPhones and Apple Watches. It also opens a credible second front in its competition with SpaceX, which has been expanding from LEO broadband into direct-to-device services with characteristic speed.
What Amazon Is Buying — And What It Still Can’t Buy
Globalstar has operated in the mobile satellite services business for decades, providing L-band connectivity that Apple adopted for satellite features on its devices. The deal preserves that Apple relationship while layering on Amazon’s plans to deploy a more advanced direct-to-device network by 2028, one that would support voice and data, not just emergency text messaging.
Apple has indicated it will continue working with Amazon on the satellite infrastructure, building on the companies’ existing collaboration through Amazon’s cloud services. That relationship alone justifies a significant portion of the deal’s value: Apple’s iPhone installed base represents the single largest addressable market for satellite-to-phone services.
Amazon also committed to supporting Globalstar’s next-generation constellation plans, including replenishment of its L-band network and a D2D expansion involving satellites being built by MDA. The deal structure reportedly offers Globalstar shareholders a mix of cash and Amazon stock. A significant portion of Globalstar’s shareholders have already approved the transaction.
Closing is expected in 2027, pending regulatory approvals and conditions tied to progress on Globalstar’s satellite replacement program. An Amazon spokesperson indicated the company will be working cooperatively with various regulators across numerous jurisdictions, seeking telecommunications license change of control approvals in the U.S. and other countries worldwide to close the transaction.
What the deal doesn’t buy is time. By 2028, when Amazon plans to launch its D2D system, SpaceX’s direct-to-device capabilities will have had four years of operational refinement, iterating on hardware already in orbit. Amazon will be entering with a first-generation product against a competitor on its fourth or fifth.
The Spectrum Moat vs. The Execution Gap
MSS spectrum is a finite resource, and Globalstar’s global licenses represent a regulatory moat that would take years to assemble through applications and auctions. SpaceX has been pursuing its own multibillion-dollar push for D2D frequencies alongside Starlink, making spectrum control an increasingly contested asset. On paper, Amazon just leapfrogged years of regulatory work with a single transaction.
But spectrum without satellites is a license to broadcast silence. And here, the contrast between the two companies is stark. SpaceX and T-Mobile have already demonstrated text services through Starlink satellites equipped with direct-to-cell antennas — hardware that was designed, built, launched, and tested while Amazon was still negotiating this acquisition. Amazon has said its D2D system will support voice and data beyond emergency messaging, but the company has not disclosed target throughput, latency benchmarks, or how its L-band architecture would compare to SpaceX’s approach of repurposing cellular band frequencies from orbit. Until those specifics materialize, the claim remains aspirational.
SpaceX’s advantage isn’t just technical — it’s iterative. The company has demonstrated an ability to update satellite hardware between launches, effectively upgrading its constellation in real time. Amazon has yet to prove it can match that cadence, and the Globalstar acquisition doesn’t change the underlying manufacturing and launch dynamics that will determine whether it can.
Amazon’s Constellation Problem
The Globalstar deal arrives at a moment when Amazon’s core satellite program is under real pressure — which is precisely why the acquisition reads as much like a hedge as a strategic advance. A limited number of satellites have been launched since deployments began, with hundreds still required to meet FCC license milestones.
Amazon has faced challenges meeting deployment deadlines and reportedly requested relief from the FCC, citing delays with key launch vehicles including Arianespace’s Ariane 6, Blue Origin’s New Glenn, and ULA’s Vulcan Centaur — three rockets that have themselves experienced development setbacks. Amazon has indicated that deploying a next-generation D2D system won’t impact its timeline or plans for its core LEO broadband service, which remains on track to roll out later this year.
Amazon LEO aims to provide initial broadband services by mid-summer. But the company’s launch cadence needs to accelerate dramatically to satisfy its FCC commitments, and betting on three rockets that are all still proving themselves is a vulnerability SpaceX simply doesn’t share.
SpaceX, by contrast, has amassed thousands of satellites in LEO, making Starlink by far the world’s largest constellation. That numerical asymmetry isn’t just a scoreboard — it represents thousands of iterations of manufacturing, launch, deployment, and on-orbit operation. It’s institutional knowledge encoded in hardware, and no acquisition can replicate it overnight.
The Regulatory Knife Fight
If the orbital gap weren’t enough, the two companies are also locked in regulatory combat at the FCC — a theater where Amazon’s deep pockets may actually match SpaceX’s advantages. SpaceX recently accused Amazon and its launch partner Arianespace of violating orbital debris requirements by launching satellites into initial altitudes higher than those specified in Amazon’s FCC filings. SpaceX alleged that Amazon was deploying satellites at unauthorized altitudes without adequate coordination or information sharing.
Amazon fired back, saying SpaceX itself helped launch Amazon satellites into similar altitudes when it served as a launch partner, and that SpaceX only objected after moving Starlink satellites into nearby orbits. Amazon also said it had informed the FCC of the launch altitudes SpaceX now contests. The spat reflects a pattern both companies have acknowledged: using FCC proceedings as a competitive weapon to slow each other down.
FCC Chairman Brendan Carr has weighed in on disputes between the two companies regarding satellite deployment plans. The political dynamics around FCC decision-making add another variable to Amazon’s regulatory timeline for both its constellation license and the Globalstar acquisition — and give SpaceX additional levers to complicate Amazon’s path to orbit.
The AWS Precedent — And Its Limits
Amazon has demonstrated patience with long-horizon infrastructure bets before. AWS was a money-losing curiosity before it became the most profitable division in Amazon’s portfolio. The company has committed billions to its satellite constellation program even before this acquisition. The Globalstar deal represents a substantial additional space investment on top of that.
The AWS analogy is the one Amazon would prefer investors to focus on. But building and operating a global satellite constellation is a fundamentally different engineering problem than building data centers on the ground. Data centers don’t deorbit when something goes wrong. They don’t require securing launch windows on rockets that may or may not be ready. And when AWS was scaling up, it wasn’t racing a competitor that had already captured the market with a five-year head start and vertically integrated launch capability.
The real strategic logic of the Globalstar deal may be less about winning the D2D race outright and more about ensuring Amazon doesn’t lose it by default. Owning spectrum and ground infrastructure gives Amazon optionality even if its constellation deployment stumbles. It preserves a path to relevance in satellite-to-phone services that would otherwise close as SpaceX and its carrier partners lock up the market.
The Bigger Picture for Commercial Space
A major space acquisition by one of the world’s largest companies sends a clear signal about where the commercial space industry is heading. Direct-to-device connectivity is moving from a niche emergency feature to a core capability that the biggest technology companies in the world consider worth billions to control.
The deal also signals consolidation. Globalstar operated for years as an independent satellite company serving specialized markets. It now becomes a component of Amazon’s space strategy. As D2D competition intensifies, other small and mid-sized satellite operators with useful spectrum or orbital assets may find themselves acquisition targets — particularly those that could give other tech giants their own path into the market before the window closes.
Amazon’s bet is that satellite-to-phone connectivity will become as foundational as Wi-Fi or cellular coverage, a utility that billions of people expect to work everywhere. If that bet is right, owning the spectrum, the satellites, and the ground infrastructure to deliver it will be worth far more than the acquisition price. If it’s wrong, or if execution falters, Amazon will have made one of the most expensive space investments in commercial history with little to show for it.
The deal won’t close until 2027. Between now and then, Amazon has to prove it can get satellites into orbit at the pace its FCC license demands, defend its orbital parameters against SpaceX’s regulatory challenges, and build a D2D network capable of competing with a rival that is already serving customers. Amazon just wrote the biggest check in the satellite-to-phone race. SpaceX’s response will be the same as it’s always been: launch more rockets, ship more hardware, and let the scoreboard speak for itself.
Photo by Chris Lyo on Pexels