Amazon has filed with the U.S. Federal Communications Commission for authority to deploy a constellation of up to 5,105 satellites designed for direct-to-device connectivity. The system would operate in Globalstar spectrum bands and is intended to deliver communications straight to unmodified smartphones and Internet-of-Things devices. Service is expected to begin around 2028, complementing Amazon’s broader Leo broadband constellation and the Globalstar network the company is acquiring. While Starlink’s aviation partnerships have dominated recent discussion of in-flight Wi-Fi, Amazon’s direct-to-device architecture represents a distinct and potentially more disruptive path.
Traditional satellite connectivity for aircraft relies on specialised antennas installed on the fuselage. Airlines must retrofit or line-fit these systems, negotiate bandwidth, contracts and manage the associated weight, drag and certification costs. Starlink and Amazon’s own Kuiper/Leo broadband offerings follow this model. Direct-to-device technology aims to bypass the aircraft antenna entirely. Passenger phones and other devices would communicate straight to the satellite, much as they already do with terrestrial cellular networks in coverage areas. If the technology matures to support reliable broadband data at aviation altitudes and speeds, the economic and operational logic of installing dedicated aircraft antennas begins to weaken.
Asian carriers are currently evaluating or negotiating antenna-retrofit programmes, the implications are significant. Many airlines in the region have been weighing expensive fleet-wide installations to remain competitive on long-haul and high-value routes. Those capital commitments are typically multi-year and difficult to reverse. A credible direct-to-device alternative arriving in the late 2020s could strand some of that investment or at least change the cost-benefit calculation for future aircraft. Carriers that have not yet signed large retrofit contracts may choose to wait. Those already committed may find themselves locked into a technology path that becomes less essential once passenger devices can connect independently.
The technology still faces substantial hurdles. Aviation use cases demand consistent performance during high-speed flight, across oceanic and polar routes, and inside a metal aircraft fuselage that naturally attenuates signals. Regulatory approval for the spectrum and operational rules at altitude will take time. Capacity and interference management across thousands of satellites and millions of potential devices remain complex. Amazon’s filing positions the constellation primarily for terrestrial and IoT applications, with aviation as a possible future extension rather than the initial focus. Nevertheless the architectural direction is clear.
If direct-to-device connectivity reaches the performance threshold required for passenger broadband, the current generation of aircraft antenna programmes could look like a transitional technology. Airlines would still need solutions for cockpit communications, operational data and premium cabin services that demand guaranteed bandwidth. But the mass-market passenger Wi-Fi business case would shift from airline-provided infrastructure to device-native satellite access. For Asian carriers investing heavily in the current retrofit cycle the timing creates a strategic tension: commit capital now to meet near-term passenger expectations, or pace investment in anticipation of a more fundamental change in how connectivity reaches the cabin.
Amazon’s 5,105-satellite application is not an immediate threat to existing aviation satellite contracts. It is an early signal that the technical foundation for a different model is being laid. In a region where airlines are still deciding how much to spend on today’s antenna systems, that signal is worth watching closely. The next frontier for in-flight connectivity may not be a better antenna. It may be no antenna at all.