On 27 October 2026 the FCC's Space Bureau, with the Office of Engineering and Technology and the Office of International Affairs, granted Amazon's application to take over Globalstar. The deal includes about 25 MHz of internationally harmonized mobile-satellite spectrum. The order cites a $11.57 billion transaction, and the FCC's own press release rounds it to $11.5 billion. Globalstar says it expects closing in 2027, subject to remaining approvals and satellite milestones.
The approval is the right call on the record. It also leaves one competition question open: whether a bigger third entrant in direct-to-device (D2D) satellite service will actually discipline the market.
The strongest case against the deal
Critics have a serious argument. Satellite spectrum is scarce, and the D2D market is forming right now. Starlink has more than 11,000 satellites, and the FCC has authorized a 15,000-satellite D2D network for SpaceX and a 248-satellite system for AST SpaceMobile, according to The Next Web. A reviewer worried about concentration could see Amazon buying a scarce L-band asset as the kind of move that locks up inputs in a nascent market. Globalstar's biggest customer is also Apple. The FCC order says Apple accounted for about 63% of Globalstar's 2025 revenue and holds a 20% equity stake that Amazon says it will acquire. A single dominant customer moving from one owner to another is exactly where lock-in worries concentrate.
Why the order's competition analysis holds
The order applies the standard FCC framework. The Commission's public-interest test is "informed by, but not limited to, traditional antitrust principles", and it says the FCC's competitive review is broader than the Justice Department's. On the facts, the bureau found no horizontal overlap. It said the transaction "does not implicate either horizontal or vertical harms" because the parties "do not compete for each other's customers."
That finding is well supported. Amazon Leo is authorized for Ka-, Ku- and V-band fixed-satellite service. Its offerings are "principally broadband FSS services delivered to fixed terminals." Globalstar sells mobile-satellite voice, messaging and Internet-of-Things connectivity over 1.6/2.4 GHz spectrum, serving roughly 811,000 subscribers as of 30 June 2026. The order also notes that Amazon holds no L-band, S-band or comparable spectrum that would let a satellite reach an ordinary smartphone.
The most important finding is the counterfactual. Amazon has a pending D2D application, but the order says it "would not have access to MSS spectrum needed to operate that system absent this transaction." That makes this an entry transaction rather than a consolidation of rivals. Amazon is authorized for 3,232 low-Earth-orbit satellites in its Leo systems. The Commission has approved 7,736 Amazon satellites in total, against authorization for only 26 Globalstar satellites, nine of them subject to a debris-mitigation plan. Putting capital and manufacturing capacity behind scarce spectrum is the efficiency gain the FCC credited. It also noted that Amazon can produce up to 30 satellites per week at its Kirkland facility.
The record was thin on adversaries
The procedural record shows how little organized opposition there was. The order lists four comments, one petition to dismiss, one opposition, one reply and several other filings. Two public-interest groups, Public Knowledge/Open Technology Institute and the Information Technology and Innovation Foundation, supported approval. Both argued the deal would produce a stronger competitor to SpaceX/T-Mobile and AST SpaceMobile/AT&T. The only substantive objection came from Yippy, Inc., a former Globalstar customer. The bureau denied its petition, calling a transaction review "not the proper forum for resolving private contract disputes." It also rejected an individual filer's claim that the FCC lacked jurisdiction.
What the order does not resolve
Three issues deserve more attention than the order gives them.
- Entry is a prediction, not a fact. The FCC "credit[ed]" claims that Amazon would put the spectrum to intensive use, citing Amazon's investments to date. But Amazon's D2D licence is still pending, and the order discounts benefits that arrive in the distant future. Regulators should hold Amazon to its own timeline. Spectrum bought and left idle would be the worst outcome for competition.
- Apple's position is the real market-structure question. The order records that Amazon expects to keep providing service to Apple. Amazon is also reported to have signed an agreement with Apple to keep iPhone and Apple Watch satellite features running. An Amazon-owned network serving a customer that supplied about 63% of its revenue is not a harm in itself. It does mean the deal's competitive payoff depends on Amazon building a multi-carrier, multi-device business, as the applicants claim, rather than serving one anchor customer.
- Narrow conditions are the right design. The bureau conditioned the grant on Amazon's continued compliance with the 1999 and 2004 letters of agreement between Globalstar and the Departments of War and Justice and the FBI. It made no attempt to impose behavioral remedies on competition grounds. That restraint is correct. Conditions meant to police hypothetical conduct in a market that barely exists tend to freeze it.
A pro-innovation reading
The policy lesson is that blocking the combination would have protected a duopoly's head start more than it protected consumers. The alternative to this deal is a Globalstar that stays a small operator with 26 authorized satellites and heavy dependence on one customer. Its financial strain is visible: the company reported a $26.5 million net loss in the second quarter. Amazon's capital is the likeliest route to a third full-scale D2D competitor.
The right follow-up is to watch outcomes, not to second-guess the approval. Regulators and competition authorities should track whether Amazon opens service to multiple device makers and carriers, and whether it meets the launch milestones its own filings imply. Non-US authorities that still have to clear the transaction should apply the same evidence-based test: look for actual overlap and a credible counterfactual, and ask whether the deal adds a competitor.