Infrastructure

Amazon Filed For 5,105 Satellites It Cannot Yet Use

Amazon asked the FCC to fly 5,105 direct-to-device satellites from 2028, using spectrum it does not own yet. What that timing means for your continuity plan.

InfrastructureBy Servola Tech Desk2026-07-274 min read

AI-assisted, edited by humans. | Editor-in-Chief: Leon Soliman | Editorial standards

Amazon Filed For 5,105 Satellites It Cannot Yet Use

Key takeaways

Amazon has filed for 5,105 satellites that would beam voice and data straight to ordinary phones from 2028, but the spectrum they run on arrives with an acquisition worth about 11 billion dollars that is not expected to close until 2027, so the date to plan against is a merger approval, not a launch campaign.
If your business continuity plan currently lists satellite direct-to-device as a 2028 capability, move it to 2030 and keep the terrestrial redundancy you were preparing to retire, because a deployment that begins in 2028 across five orbital shells does not become useful coverage in the same year it starts.
Check which of your mobile contracts sits with Vodafone, because Amazon named it as a partner and this capability will almost certainly reach you as a line item on a carrier bill you already pay, not as a new supplier you get to select.

A filing number, five shells and a 2028 start

On 24 July Amazon lodged application SAT-LOA-20260601-00224 with the US Federal Communications Commission. The document asks for authority to deploy up to 5,105 low Earth orbit satellites in a system Amazon calls Leo Direct-to-Device, built to reach compatible smartphones and wearables without any dish, terminal or intermediate hardware. Deployment would begin in 2028.

The engineering detail in the filing is unusually specific for a system this far out. The constellation is arranged in five shells: 580 kilometres at 55.7 degrees inclination, 560 kilometres at 58.2 degrees, 510 kilometres at 56.3 degrees, 570 kilometres at 73 degrees, and 540 kilometres at 84 degrees, the last of these close enough to polar to cover the far north. Individual satellites are rated for a service life of six to eight years. Links to the handset use L-band and S-band; Ka-band and V-band carry the traffic down to ground stations.

Amazon is not starting from nothing. More than 390 satellites of its broadband Leo constellation are already in orbit, and the company has named Vodafone, DirecTV, Herotel in South Africa and Australia's National Broadband Network as partners. The service described in the filing covers messaging, voice, mobile data, emergency communications, in-vehicle connectivity and machine-to-machine traffic, with supported iPhone and Apple Watch models cited among target devices.

The airwaves belong to somebody else until 2027

Read the filing next to the spectrum it assumes and the timeline changes character. Amazon is not asking the regulator for a new allocation. It intends to operate at 1.6 and 2.4 gigahertz inside the United States, and 1.5 and 1.6 gigahertz elsewhere, and those are Globalstar's licences. Amazon is buying Globalstar in a transaction valued at roughly 11 billion dollars, or about 10 billion euros, and that purchase is expected to close in 2027.

This is the part that reorders the risk. A satellite programme normally slips because rockets are scarce, manufacturing is slow or a regulator asks hard questions about orbital debris. Here the binding constraint sits earlier in the chain and in a different building: an antitrust and licence-transfer review of a large acquisition. If that review runs long, or attaches conditions, the constellation does not simply launch on schedule into spectrum it can use. The FCC's April 2026 order on direct-to-device access already refused competing requests for spectrum, which is precisely why buying an incumbent licence holder was the faster route than asking for a fresh grant.

The competitive field reinforces the point rather than softening it. Rocket Lab paid 8 billion dollars for Iridium; SpaceX runs Direct to Cell on partner carrier spectrum today. What separates these programmes is not who can build satellites but who already holds, or has bought, mobile satellite spectrum that regulators will let them point at consumer handsets. The hardware is becoming the commodity. The licence is the asset.

What belongs in the continuity plan this quarter

The practical error available to an operator right now is to treat 2028 as a coverage date. It is a deployment start date for a constellation of up to 5,105 satellites whose spectrum foundation is still subject to a pending transaction. Five shells do not populate in a year. A realistic planning assumption for usable, contract-backed direct-to-device service in Europe is the early 2030s, and any redundancy you were going to decommission on the strength of a 2028 headline should stay where it is.

The second and more immediate consequence concerns procurement rather than timing. Direct-to-device is not sold to businesses as a standalone satellite service. It arrives through the mobile network operator, which is why the named partner list matters more than the satellite count. For a European operator that means the eventual commercial terms are set inside a Vodafone contract or its equivalent, negotiated by a carrier, and presented to you as an option on an existing bill. The moment to influence that is when the framework agreement is renewed, not when the service launches.

There is a narrower group for whom the filing changes the arithmetic sooner. If you run field operations, logistics fleets, offshore or remote sites, or distributed sensors in places where terrestrial coverage is genuinely absent, the relevant question is what you are paying today for specialist satellite messaging hardware on a replacement cycle that will run past 2028. Kit bought now on a five-year depreciation schedule will still be on the books when a handset-native alternative reaches the market. That is a specification decision to make deliberately, not a purchase to defer.

Frequently asked questions

Will this replace our mobile contracts in Europe?

No. Direct-to-device is designed to fill gaps in terrestrial coverage, not to substitute for it, and Amazon's own partner list shows the commercial route runs through mobile network operators rather than around them. Vodafone is the European name in that list. The realistic outcome for a European business is an additional capability offered within a carrier agreement, priced by the carrier, covering messaging and emergency communications first and broader data later. Treat it as an extension to an existing supplier relationship and negotiate it at contract renewal.

Should we delay buying satellite communication equipment for remote sites?

Only if the equipment you are replacing can safely last until the early 2030s, which is the honest window for a system that begins deploying in 2028 on spectrum tied to a transaction closing in 2027. For most operations the answer is to buy what the site needs now, but to shorten the depreciation assumption and avoid multi-year exclusivity on specialist hardware. The decision to make deliberately is contract length, not the purchase itself.

The number in the headline is 5,105, and it is the least important figure in the filing. The one that governs the schedule is the pending acquisition that carries the spectrum, and it closes a year before the first satellite of this system is due to deploy. Plan the capability for the decade it will actually arrive in, keep the redundancy you already have, and put the conversation where it will really be decided, which is your next mobile framework negotiation.

Amazon LeoGlobalstarFCCdirect-to-devicesatellite connectivityVodafonespectrumbusiness continuity

Cadence And Synopsys Tuned Their Tools For Nvidia

Nvidia says its own Vera CPU runs Cadence and Synopsys verification up to 1.5x faster. The number is Nvidia's, and the design farm question is now open.

5 min read

Infrastructure

Nvidia May Guarantee 250 Billion of OpenAI's Rent

OpenAI holds no investment-grade rating, so Nvidia is reportedly guaranteeing the lease on a 10 gigawatt Ohio campus. A chipmaker's credit is now a line in your compute budget.

4 min read

Law & Regulation

The EA Buyout Now Hinges on Brussels

The largest take-private in history has cleared EA shareholders and US antitrust. The last real gate is the EU - and not classic merger review, but the Foreign Subsidies Regulation, facing its first blockbuster test.

3 min read

← All articles