Amazon Leo filed an application with the FCC on July 24 requesting authorization for a direct-to-device satellite system that would deliver voice calls, text messages, mobile data, and emergency services directly to unmodified 4G and 5G smartphones, with service targeted no earlier than 2028. The application requests authorization for 5,105 satellites, using Globalstar's globally licensed L-band and S-band mobile satellite spectrum, and Amazon has already signed early partnership agreements with Vodafone, DirecTV, Herotel, and Australia's NBN. For IT leaders managing connectivity for distributed or remote operations, this is a filing worth tracking now, even with a 2028 target — but it's also worth being clear-eyed that Amazon is arriving into a market SpaceX already has operating at commercial scale.
The competitive gap Amazon is filing into
SpaceX's Starlink Direct to Cell service has been live commercially on T-Mobile's network since July 2025 — meaning by the time Amazon Leo's proposed network could plausibly go live in 2028, Starlink will have had roughly three years of real-world operational experience, carrier partnerships, and iterative service improvement that Amazon has yet to accumulate. That head start matters beyond simple market-share bragging rights: direct-to-device satellite connectivity is a genuinely difficult engineering problem involving spectrum coordination, handset compatibility across manufacturers and chipsets, and network densification to achieve meaningful coverage and throughput. Starlink's multi-year operational lead means it has already worked through a substantial share of the real-world engineering and regulatory friction that Amazon Leo's filing suggests it's only now formally beginning to navigate at scale.
Why the spectrum partnership choice is the more interesting detail
Amazon's application relies specifically on Globalstar's globally licensed L-band and S-band mobile satellite spectrum, rather than spectrum Amazon holds directly or has separately negotiated. That dependency is a meaningful strategic detail: it means Amazon Leo's direct-to-device ambitions are partly gated by its relationship with Globalstar and the scope of Globalstar's existing global licensing footprint, rather than purely by Amazon's own satellite manufacturing and launch cadence. For enterprise IT planners evaluating any future Amazon Leo direct-to-device offering, this dependency is worth tracking as a risk factor distinct from Amazon's general execution capability — a strong satellite constellation build-out timeline doesn't guarantee the spectrum and regulatory arrangements clear on the same schedule.
Reading the early carrier partnerships correctly
Vodafone, DirecTV, Herotel, and Australia's NBN as early direct-to-device partners gives Amazon Leo a genuinely international footprint from the outset, spanning European, US satellite TV, African, and Australian markets respectively. That's a broader initial geographic spread than Starlink's initial T-Mobile-only US launch, which could position Amazon Leo more competitively for organizations with international distributed operations if the service actually launches on schedule and delivers comparable performance. But signed early partnership agreements at the FCC filing stage are a substantially weaker signal than commercial service already running — Starlink's T-Mobile partnership converted from a similar early-stage announcement into live commercial service, and that conversion process is exactly the multi-year execution risk Amazon Leo hasn't yet cleared.
Why the Globalstar spectrum relationship is worth understanding in more depth
Globalstar has historically operated as a smaller, more specialized mobile satellite services provider compared to giants like Iridium, and its L-band and S-band spectrum licenses span a genuinely global footprint that took years of individual country-by-country regulatory negotiation to assemble. Amazon's decision to build its direct-to-device ambitions on top of Globalstar's existing licensed spectrum, rather than pursuing its own from-scratch global spectrum licensing process, is a pragmatic choice that significantly shortens Amazon's path to global coverage compared to starting that regulatory process independently — but it also means Amazon Leo's ultimate service footprint is bounded by wherever Globalstar's existing licenses actually extend, and any gaps in that global licensing coverage become gaps in Amazon's own service map unless separately addressed. Enterprise connectivity planners evaluating Amazon Leo as a future option should specifically ask, once more detail becomes available, which countries and regions are covered under the Globalstar spectrum arrangement Amazon is relying on, rather than assuming "global" coverage claims apply uniformly everywhere your organization operates.
What this means for enterprise connectivity planning
-
Don't factor Amazon Leo direct-to-device service into any connectivity planning with a horizon shorter than 2028, and treat even that date as an optimistic best case given the regulatory approval process still ahead and Amazon's lack of operational track record in this specific service category. FCC filings represent the start of a regulatory and engineering process, not a committed launch date.
-
If your organization has remote, maritime, or otherwise cellular-dead-zone operations requiring direct-to-device satellite connectivity today, Starlink Direct to Cell via T-Mobile is the only commercially available option at meaningful scale right now — evaluate it on its current merits rather than waiting to compare against a competing service that doesn't yet exist commercially.
-
Watch Amazon's FCC approval timeline as a leading indicator for the broader direct-to-device satellite market's regulatory pace, since the same spectrum coordination and interference questions the FCC will work through for Amazon's application are relevant to any future entrant in this space, not just Amazon specifically.
-
Reassess vendor diversification value in this category over the next several years, not immediately. If Amazon Leo does launch with genuinely international carrier partnerships already in place, it could offer a meaningfully different geographic coverage profile than Starlink's current footprint — worth tracking as a future option for global connectivity strategy, even though it shouldn't influence procurement decisions today.
-
Treat this filing as one more data point in Amazon's broader infrastructure investment pattern, alongside its reported $25 billion bond sale to fund AWS AI infrastructure and its on-device AI chip development. Amazon is making large, parallel infrastructure bets across cloud compute, custom silicon, and now satellite connectivity — useful context for understanding Amazon's long-term platform strategy even if this specific satellite filing doesn't affect your procurement decisions for years.
What the FCC approval process actually involves
Filing an application is the beginning of a regulatory process that involves substantially more scrutiny than a simple rubber-stamp approval, particularly for a constellation of this size. The FCC will need to evaluate orbital debris mitigation plans, coordinate spectrum usage to avoid harmful interference with Globalstar's existing licensed operations and other satellite operators using adjacent spectrum, and assess the technical feasibility of Amazon's proposed satellite-to-phone architecture at the scale proposed. Given that Starlink's own Direct to Cell buildout went through a comparable regulatory process before reaching commercial launch, Amazon Leo's timeline is a reasonable point of comparison — but Amazon is also proposing a meaningfully larger satellite count than Starlink's initial Direct to Cell deployment, which could mean either a longer regulatory review given the increased scale of orbital and spectrum coordination involved, or a smoother process if Amazon has structured the filing to address concerns regulators already raised during Starlink's comparable review. Either way, the 2028 target date in Amazon's own filing should be read as the optimistic end of a range, not a firm commitment, given how much regulatory and engineering work remains between filing and commercial service.
How this fits Amazon's broader satellite ambitions
Amazon Leo direct-to-device is a distinct initiative from Amazon's existing Project Kuiper satellite internet service, which targets fixed and mobile broadband terminals rather than unmodified smartphones, and it's worth keeping the two efforts conceptually separate when evaluating Amazon's overall satellite strategy. Kuiper has already begun commercial rollout in select markets, including plans for satellite internet service in Pakistan and other regions during 2026, giving Amazon operational satellite constellation experience that could meaningfully de-risk the direct-to-device buildout compared to if Amazon were attempting both efforts as a first-time satellite operator simultaneously. For enterprise customers already evaluating or using Kuiper for fixed broadband connectivity at remote sites, Amazon's growing satellite portfolio is worth tracking as a single vendor relationship that may eventually offer both fixed broadband and direct-to-device mobile connectivity from one commercial relationship — a potential procurement simplification, several years out, worth flagging to your vendor management team as a future consideration even though it shouldn't drive any near-term decision.
The bigger picture: direct-to-device is becoming genuinely competitive
Regardless of how the Amazon-versus-Starlink race plays out, the broader significance here is that direct-to-cell satellite connectivity — the ability for an unmodified consumer smartphone to connect directly to a satellite without special hardware — has moved from a niche emergency-service feature to an active, multi-player competitive market within the space of about a year. That's genuinely useful context for any enterprise connectivity strategy: coverage gaps that used to be permanent constraints for field operations, logistics, or remote facility management are increasingly likely to have a satellite-based fallback option within a few years, even if today's practical options remain limited to Starlink's existing service. The organizations that benefit most from this shift will be the ones tracking the competitive landscape now, so they're positioned to evaluate genuinely competing options as they become commercially available, rather than defaulting to whichever provider happened to launch first.