Amazon's $9B regulatory insurance policy: the Globalstar spectrum play

Amazon, Apple, and SpaceX are all converging on the same asset. It can't be built, it can't be duplicated, and there's a finite amount of it on or above Earth. Amazon is in talks to acquire Globalstar for roughly $9 billion. Globalstar operates a small LEO satellite constellation with plans to expand. While the fleet is relatively small today, the spectrum portfolio behind it... well that's anything but. Globalstar holds licensed orbital spectrum in the S-band, already cleared for 5G use across dozens of regulated markets worldwide. This kind of spectrum can't be manufactured. It sits at the intersection of physics, international law, and national licensing regimes. Once allocated, no amount of capital expenditure can replicate it. Meanwhile, Amazon's Kuiper constellation faces real deployment pressure. The company has already formally asked the FCC for a two-year extension on its July 2026 deadline to have half its constellation in orbit. Acquiring Globalstar gives Amazon licensed frequencies, existing ground infrastructure, and a faster path into direct-to-device connectivity. A meaningful shortcut versus building that position from scratch. One analyst called it a \"$9 billion regulatory insurance policy.\" The deal is far from clean. Apple has invested $1.5 billion in Globalstar, holds a 20% equity stake, and currently reserves 85% of network capacity for iPhone emergency and messaging services. Any acquisition runs through Apple's contractual and economic position first. SpaceX was also reportedly evaluating a bid as of late 2025. The satellite industry focuses on launch costs, constellation size, and orbital coverage. Those matter. But without licensed frequencies, a constellation is hardware with no right to transmit. Spectrum is the layer beneath all of it. Launch costs are falling. Satellites are getting cheaper to build. But spectrum... it isn't getting cheaper at all.