AST’s $1 Billion Lesson in Satellite Scale

In January 2025, Vodafone used an ordinary smartphone to make a video call through AST SpaceMobile satellites and said European service would begin later that year. In 2026, after Blue Origin’s New Glenn failed to place a BlueBird satellite in its intended orbit, AST pushed service to 2027 and planned a $1 billion convertible-notes raise. The phone connected; the network slipped two years.

AST proved the radio link before it built a commercial network. Direct-to-device competition now rewards the operator that can keep orbital capacity launched, licensed, financed and sold through a rocket failure or delayed transaction. AST relies on carriers for distribution and spectrum access, SpaceX controls more of launch, and Amazon is buying or coordinating missing layers around Leo.

The handset stopped being the moat

AST began with the right technical diagnosis. In 2020, then called AST & Science, the company raised money to build a broadband constellation that would link directly to smartphones rather than require proprietary terminals. NASA warned that AST’s proposed satellites posed collision risks, an early sign that ordinary-device coverage would make fleet operations as consequential as radio engineering.

Handset technology then moved toward AST’s premise. In 2023, Samsung announced a standardized 5G non-terrestrial-network modem for two-way communication between smartphones and satellites. Device makers were turning satellite compatibility from a bespoke feature into a reusable communications block.

Vodafone supplied the most visible proof in January 2025 with its satellite video call through AST. The later service delay does not erase that demonstration; it defines its limit. AST showed that the radio link worked, but not that enough capacity would remain continuously available across a commercial footprint.

To deliver dependable mobile service, an operator must coordinate satellite count, beam capacity, ground stations, terrestrial backhaul, spectrum rights and replacement launches. SpaceX President Gwynne Shotwell made the terrestrial requirement explicit when she said the company aims to add ground infrastructure for a “true mobile service.” Larger satellites add capability but concentrate more capacity on each launch. More launch providers reduce supplier dependence but add integration work. Carrier spectrum speeds market access while preserving carrier leverage.

A carrier choosing a satellite partner now buys replenishment capacity, regulatory access and operational continuity along with the radio link.

Carriers opened markets but could not secure orbit

AST secured its strongest commercial layer before it finished its constellation. AT&T signed an agreement running through 2030. Verizon committed $65 million in prepayments and $35 million in debt alongside its direct-to-phone broadband partnership. Vodafone later joined AST in planning a Europe-led constellation operated from Germany.

AT&T, Verizon and Vodafone bring subscriber relationships, billing systems, licensed cellular positions and experience securing national approvals. Vodafone’s Germany plan also gives European governments a regional operating structure instead of a purely foreign satellite overlay. AST can enter markets through institutions that already know how to sell and regulate mobile service.

Those partners also spare AST from building three national retail operations before selling capacity. Verizon’s prepayment converts part of future demand into present deployment support, while its debt commitment shares some financing risk.

Carrier contracts cannot place a satellite into its intended orbit, manufacture replacement capacity or guarantee a launch slot. AST aligned sales, spectrum pathways and local access, but launch procurement and fleet replenishment still operate on separate failure clocks.

One missed orbit forced another financing cycle

Low-Earth-orbit networks consume launches continuously. Operators must deploy an initial fleet, fill coverage gaps, replace failed spacecraft and refresh aging capacity. They need a launch cadence that can absorb delays without pushing service into another financing cycle.

AST’s largest-ever satellite launch in December 2025, carrying BlueBird Block-2, was presented as the first in a series of deployments. AST needed the series more than the individual spacecraft: one successful launch creates an asset, while repeated launches create a network.

Blue Origin’s New Glenn subsequently failed to place an AST BlueBird satellite in its intended orbit. AST shares fell more than 7% in early trading, while the operational consequences ran deeper: AST delayed service to 2027 and planned a $1 billion convertible-notes raise.

The gap between AST’s two financing milestones captures the economic distance between proving a direct smartphone link and sustaining a fleet that can sell it as a service.

A constellation operator keeps capital available to reorder hardware, procure another launch, carry operating costs through delays and continue ground-network work before revenue arrives. Without that reserve, a launch failure becomes a service delay, which then becomes a financing event under worse conditions.

Blue Origin’s New Glenn—not SpaceX—failed to place AST’s satellite, and the record does not establish that AST buys SpaceX launches. AST’s exposure is dependence on outside launch capacity itself. SpaceX still gains from controlling a major source of that capacity while rivals face production and launch bottlenecks.

AST can spread the risk across launch providers, hold larger capital buffers and avoid concentrating too much commercial capacity on one mission. Redundant procurement raises coordination costs, cash reserves reduce capital efficiency, and smaller launch batches can slow coverage buildout. Operators pay for those options because rocket failures make them valuable.

SpaceX and Amazon remove different failure points

SpaceX entered direct-to-cell service from a different position. It manufactures Starlink satellites, operates the constellation and launches its own spacecraft. SpaceX deployed its first six direct-to-cell-capable Starlink satellites in January 2024 with T-Mobile, then pursued terrestrial infrastructure to complement the orbital network. The company can coordinate a satellite revision with a launch manifest inside one operating system.

SpaceX has also added licensed spectrum. In May 2026, the FCC approved EchoStar’s sale of approximately 65 MHz to SpaceX and 50 MHz to AT&T. The transaction gives both a satellite operator and a terrestrial carrier more control over a resource that regulators must convert into usable hybrid-network capacity. That conversion belongs to the broader market for satellite spectrum reallocation, where institutional clearance can matter as much as radio performance.

Amazon is assembling a different system. In April 2026, Amazon agreed to acquire Globalstar for about $11.6 billion in a transaction expected to close in 2027. The deal includes Apple’s roughly 20% Globalstar stake, and Amazon and Apple said Leo would power some iPhone and Watch services. Globalstar gives Amazon an operating satellite platform, spectrum position, authorizations and an established device relationship.

Amazon also said its existing Leo fleet could support continuous service across initial latitudes, while the FCC approved a much larger second generation. That orbital scale makes Amazon a credible counterweight to SpaceX.

Amazon still depends on outside launches. The company asked the FCC for a two-year deployment extension while citing a rocket shortage, and it has contracted with providers including Blue Origin. Amazon draws its advantage from acquisition capital, infrastructure spending, device relationships and an approved constellation, while contracted rockets remain a constraint.

When one layer fails, AST can turn to carrier partners and capital markets. SpaceX can alter production and launch scheduling internally. Amazon can deploy acquisition capital and coordinate several infrastructure businesses, but it still needs contracted rockets to arrive. Each operator pays a different price for the same failure.

Spectrum lets carriers keep leverage

Satellite operators cannot order spectrum like another component. Regulators assign, clear and authorize frequencies, while national carriers already hold licenses and customer relationships built around terrestrial mobile service. AT&T’s approved purchase of approximately 50 MHz from EchoStar shows why AST’s carrier strategy can retain value as vertically coordinated rivals expand.

Vodafone’s proposed Germany-based constellation turns that regulatory position into a sovereignty argument. European customers would receive satellite-to-smartphone service through a regional operator and operating base. Vodafone can offer governments a local institutional counterweight to a network controlled entirely by a US satellite platform.

AT&T, T-Mobile and Verizon also signed an agreement in principle in 2026 to form a joint venture aimed at ending US wireless dead zones. The carriers have incentives to coordinate before a satellite partner gains enough spectrum, capacity and consumer access to dictate wholesale terms. Their customers want seamless coverage; their shareholders would prefer to keep control of the relationship.

Regulators decide which companies can combine terrestrial licenses with orbital coverage, which national markets require local operating structures and how much spectrum can move from legacy satellite use into mobile service. A technically superior satellite cannot route around those decisions.

Standalone service changes who owns the customer

T-Mobile and Starlink have already tested how far a carrier partnership can stretch. In July 2025, T-Satellite became available as a standalone US subscription beyond T-Mobile’s own customer base. Users could add the satellite-powered service without moving their primary mobile plan to T-Mobile.

The initial product supported texting and location sharing rather than broadband-class mobile service. A broadly available subscription does not prove that satellites can replace terrestrial capacity for video, dense urban traffic or ordinary high-volume data use. Terrestrial carriers retain towers, licensed spectrum, billing relationships and the capacity required where customers generate most traffic.

SpaceX and Charter have nevertheless held executive-level discussions about a consumer mobile offering that could give SpaceX a more direct route to phone users. Such a route would change wholesale negotiations even if satellite service remained concentrated at the coverage edge. A supplier with an alternative retail channel can bargain differently from one that depends entirely on carrier resale.

AST’s alliances spare it from building that channel now, but they also leave carriers between AST and the customer. SpaceX can use T-Mobile while testing paths beyond T-Mobile’s subscriber base. Amazon can connect Leo to Globalstar’s authorizations and Apple devices if its acquisition closes as planned. Those arrangements assign customer ownership differently, determining who captures the value as emergency messaging develops into routine data service.

Vodafone’s 2025 call proved that an ordinary phone could reach orbit. AST’s 2026 delay showed what the call could not: whether an operator can replace the next unit of capacity without reopening launch, spectrum, financing and distribution. A demonstration proves the radio; replenishment proves the network.