The satellite connectivity market woke up to the news that Amazon is in advanced talks to acquire Globalstar, the low-earth-orbit satellite telecommunications group, as the tech giant accelerates its push to build a commercial LEO satellite business capable of rivalling SpaceX’s Starlink.

The news sent Globalstar’s already elevated shares up a further 24% in after-hours trading, a striking reaction for a company whose stock had already more than doubled over the previous 12 months.

The potential deal is estimated at up to $9 billion. For context, that is roughly the same figure Amazon spent building its first 200-plus production satellites for Amazon Leo.

AWS and Amazon Leo have been designed from the outset to operate as a vertically integrated stack; owning Globalstar’s spectrum and ground station network would extend that integration considerably further.

Why Globalstar, and more importantly why now

Globalstar’s shares have surged partly on the strength of its valuable L-band spectrum and its existing satellite and ground infrastructure. Spectrum, particularly in the L-band, is a finite and jealously guarded resource. Acquiring it via a corporate deal rather than an FCC auction is a far more efficient route for a company on a tight deployment timeline.

Amazon Leo (formerly known as Project Kuiper before a rebrand in late 2025) is working towards deploying more than 3,200 satellites, with a regulatory requirement to have half its first-generation constellation in orbit by mid-2026.

With around 212 production satellites launched as of December 2025, the pace of deployment needs to accelerate dramatically. Absorbing Globalstar’s existing 24-satellite LEO constellation, ground station network spanning 24 global gateways, and its licensed spectrum across more than 120 countries would provide an immediate operational foundation that years of internal development could not replicate quickly enough.

Amazon’s broadband satellite constellation currently numbers more than 200 satellites in orbit, while Starlink operates more than 10,000. The gap is not something spectrum deals alone can bridge, but Globalstar provides Amazon with several things that launching more Ka-band satellites cannot: L-band and S-band diversity, decades of operational expertise, and infrastructure that already serves voice, data, and asset-tracking customers across enterprise and government markets globally.

The Apple complication

Any straightforward narrative around this deal runs into one significant complication: Apple. Apple acquired a 20% stake in Globalstar in late 2024, committing $1.5 billion including upfront infrastructure prepayments to help build out Globalstar’s satellite network, and that stake has made Amazon’s negotiations considerably more complex.

The depth of Apple’s operational reliance on Globalstar makes this more than just a financial stakeholder situation. Globalstar reserves 85% of its network capacity for Apple’s Emergency SOS via Satellite service, which has been built into every iPhone 14 and later model.

That service uses Globalstar’s L-band and S-band spectrum, with emergency messages routed via Globalstar’s ground stations to emergency services worldwide. Amazon, in other words, would be acquiring the critical safety infrastructure that supports hundreds of millions of active iPhones globally. Negotiating a path through that dynamic, two competing technology giants sharing infrastructure on which consumers depend in emergencies, is not straightforward.

For Amazon to successfully acquire Globalstar, it must reach a workable compromise with Apple over infrastructure sharing and future technology roadmaps, an unprecedented situation in the tech industry.

We reported earlier this year that Jeff Bezos’ space company, Blue Origin, had requested permission from the US government to launch a network of 51,600 data centre satellites.

However, the company is not the first to launch operations in space, with companies like SpaceX , Google and Amazon already deploying satellites in space for global internet coverage.

This follows comments made by the Amazon founder predicting gigawatt-scale data centres within two decades to solve soaring energy demands.