Amazon Leo Wants 5,105 Satellites To Take Phone Coverage Beyond Cell Towers

Amazon is no longer treating satellite internet as only a broadband alternative for homes, ships, enterprises and remote communities. The company now wants Amazon Leo to talk directly to ordinary mobile devices, a move that would put it deeper into the same direct-to-phone race that SpaceX, Apple, Globalstar and mobile operators are trying to shape.

In an official explanation of the plan, Amazon said Leo has filed an application with the U.S. Federal Communications Commission to launch and operate a Direct-to-Device system made up of as many as 5,105 low Earth orbit satellites. The goal is to support voice, messaging, data and emergency services on compatible mobile devices, with deployment expected to begin in 2028 if regulators approve the plan.

That is a big expansion of Amazon Leo beyond its original broadband satellite network. The current Leo constellation is designed to deliver internet through user terminals, in the same general market where Starlink has already built a large early lead. Direct-to-device service is different. It tries to reach phones without requiring a separate dish or special ground terminal, which makes it much more relevant for emergency communications, rural coverage, transport, disaster response and basic connectivity in places where towers are too expensive to build.

The Africa angle is obvious. Mobile coverage has improved across the continent, but many rural and low-income communities are still either poorly served or completely outside reliable broadband reach. Direct-to-device satellite service could help fill gaps where geography, power, roads and weak economics make tower rollout difficult. It will not replace terrestrial mobile networks, but it could become a useful layer for emergency services, messaging, IoT devices and basic internet access in hard-to-reach areas.

Amazon says the system is being developed to work with mobile network operators rather than bypass them. That distinction matters. A satellite network that partners with telcos can extend existing mobile services instead of forcing users to sign up for a separate satellite subscription. For African operators, this could eventually mean wider coverage without building every last tower themselves, although pricing, handset compatibility and local licensing will decide whether the technology becomes practical or remains premium.

The technology itself is also becoming more sophisticated. Amazon says the D2D satellites will process signals in orbit before relaying them, while using advanced signal-processing tools to improve spectrum efficiency and reduce interference. That matters because connecting directly to small mobile devices is much harder than connecting to a larger satellite terminal. Phones have limited power, small antennas and no room for a customer to adjust equipment like a dish.

This move also connects Amazon more closely with Apple. Amazon agreed earlier this year to acquire Globalstar assets and support satellite services for iPhone and Apple Watch users, while Apple continues expanding emergency and messaging features for devices outside normal coverage. Apple has already shown that satellite connectivity can become a consumer safety feature. Amazon wants to turn that into a broader network layer.

The timing fits Amazon wider AI and cloud infrastructure story. The company just reported strong AWS growth, but also showed how expensive infrastructure has become as AI, cloud and connectivity spending rises. Its latest Amazon earnings showed AWS demand accelerating even as free cash flow came under pressure from AI-related capital spending. Satellite networks add another heavy infrastructure bill, but they could also give Amazon a long-term communications platform that links cloud, devices and remote connectivity.

There are still serious hurdles. Regulators have to approve the constellation, spectrum coordination will be difficult, and Amazon has to launch thousands of satellites while avoiding delays, debris concerns and interference disputes. SpaceX already has enormous scale in orbit, and Starlink keeps expanding even with technical setbacks. Amazon is chasing a moving target.

Even so, the direction is clear. The next phase of mobile coverage may not be built only from towers. It may be a layered system where terrestrial networks handle most traffic, while satellite networks cover the gaps, disasters and remote operations. If Amazon can execute, Leo could become more than a Starlink rival. It could become part of the future mobile network stack.