The drone war's nervous system is privately owned. Ukraine's command-and-control backbone runs on tens of thousands of Starlink terminals rented from a single American company, and Russia's parallel network runs on the same constellation via smuggled terminals. Between 2026 and 2030 the reckoning arrives: Starshield's militarization, Amazon Kuiper entering service, Europe's IRIS² sovereignty push — and the birth of "C2 sovereignty" as a formal procurement category. Every military planning drone operations now has to answer a question no staff college taught in 2021: whose satellites, and whose terms?
The drone war's nervous system is privately owned. Ukraine's command-and-control backbone runs on tens of thousands of Starlink terminals rented from a single American company, and Russia's parallel network runs on the same constellation via smuggled terminals. Between 2026 and 2030 the reckoning arrives: Starshield's militarization, Amazon Kuiper entering service, Europe's IRIS² sovereignty push — and the birth of "C2 sovereignty" as a formal procurement category. Every military planning drone operations now has to answer a question no staff college taught in 2021: whose satellites, and whose terms?
The Backbone Nobody in Kyiv Owns
The dependence began as salvation. Starlink terminals flooded into Ukraine in the war's first weeks of 2022 — donated, crowdfunded, and purchased by allies — and passed 42,000 units by mid-2023, becoming the connective tissue of everything from artillery fire-control apps to drone team coordination. Then came the politics. In September 2022, according to reporting confirmed in 2023, Elon Musk declined a Ukrainian request to activate service over Crimea for a naval drone attack — a targeting decision made in a corporate headquarters, not a general staff. There were disputes over who pays, disputes over terms of service, disputes over geofencing. The Pentagon eventually regularized the arrangement with a contract in 2023, but the lesson had already landed in every Ukrainian operations room: the link works until someone in Hawthorne decides it should not.
This is not an argument that SpaceX behaved badly. It is an observation that a private company's customer-service decisions acquired the weight of allied policy, and that no treaty, no NATO working group, and no procurement regulation had anticipated the arrangement.
Moscow's Smuggled Sky
Russia's official position is that Starlink is banned. Russia's operational position is that it is indispensable. Ukrainian intelligence has documented Russian forces using thousands of gray-market terminals — purchased through intermediaries in third countries, activated under false registrations, and integrated into assault-group communications and drone C2. SpaceX says it deactivates terminals it can identify as unauthorized; the cat-and-mouse continues, with Kyiv formally requesting deactivation lists and Moscow's quartermasters finding new supply routes. Both sides of the largest drone war in history ride the same private constellation. One side rents it. The other shoplifts it.
The seizure ledger tells the scale. Captured terminals turn up in every overrun Russian position, serial numbers traceable through activation records, and each one confirmed in use is a datapoint in the argument that export-control regimes designed for missile components were never built for consumer electronics that double as theater-level C2. A satellite terminal is now controlled military technology that fits in a carry-on and ships by parcel post.
Russia's own alternatives remain years behind. The Sphere constellation program has slipped repeatedly, and existing Russian systems like Gonets offer messaging-grade bandwidth unsuitable for drone video. Moscow's answer to its LEO gap has been, in large part, theft-adjacent dependence on the adversary's ally's network — a sentence that would have been rejected as implausible in any 2019 war-game scenario.
The Constellation Race Gets a Uniform
The military market noticed, and the industry responded. Starshield, SpaceX's government arm, absorbed the defense business — including a reported $1.8 billion National Reconnaissance Office contract disclosed by Reuters in 2024 — and now builds what Starlink improvised: proliferated LEO as deliberate military infrastructure. The U.S. Space Development Agency's transport layer, built by multiple vendors, extends the model. Amazon Kuiper launched its first operational satellites in April 2025 and enters service through 2026; military offerings follow commercial ones as surely as night follows day, and every procurement office in NATO has already requested the briefing. Europe's answer is sovereignty-branded: IRIS², the €10.6 billion constellation concession signed in December 2024 with the SpaceRISE consortium, exists explicitly so that European forces never face a Hawthorne problem.
Ukraine, meanwhile, is doing the pragmatic thing: terminal diversity. Starlink remains the backbone, but Kyiv's planners now treat single-vendor C2 as a vulnerability class in its own right — mixing constellations, layering in military satcom, and pushing mesh and fiber relays to the tactical edge so that no single policy decision in a foreign boardroom can silence a front. It is the architecture of a country that learned the hard way what a kill-switch feels like, and intends never to feel it again.
C2 Sovereignty Enters the Procurement Form
By 2027 the concept has a name in acquisition documents: C2 sovereignty — the requirement that a state's command links survive the commercial and political decisions of their owners. Expect it to appear in tenders the way "ITAR-free" once did, as a formal evaluation criterion with budget lines attached. Expect GNSS hardening to ride alongside it, because the other half of the satellite problem is navigation: jamming and spoofing from the Baltic to the Black Sea through 2024–2026 has already made GNSS-denied operation a design requirement, and commercial drone insurers are beginning to write denied-environment clauses into coverage. The constellation you rent and the signals you trust are now the same risk register.
The first question of drone warfare used to be "what can it hit?" Now it is "whose satellites, whose terms — and who can switch it off?"
What to Watch
- Starshield contract expansions. New government awards reveal how fast proliferated LEO converts from improvisation into formal military infrastructure. - Kuiper military service entry. Amazon's first defense-facing offering is the duopoly moment — the first time any military gets a real second vendor for LEO broadband C2. - IRIS² funding milestones. Whether Europe's sovereignty constellation hits its schedule and budget decides whether "C2 sovereignty" is a capability or a communiqué. - Starlink policy incidents. Every geofencing decision, deactivation list, and terms-of-service change is now a geopolitical event with operational consequences. - GNSS-denial insurance requirements. Commercial insurers pricing drone operations around jammed airspace signals that navigation warfare has become a permanent environmental condition.
By 2030 no serious military will field a drone force without an owned-or-treatied answer to its command links — the era of running a war on someone else's satellites, and someone else's terms, is already closing.
The drone war's nervous system is privately owned. Ukraine's command-and-control backbone runs on tens of thousands of Starlink terminals rented from a single American company, and Russia's par…