For decades, military space was the business of a few select nations. Reconnaissance and secure communications rested on a handful of exquisite, school-bus-sized satellites that had exquisite capabilities but likewise exquisite prices, in assets so scarce that the loss of a single one could blind an entire program. The war in Ukraine has accelerated the unravelling of that model. A country with virtually no satellites of its own has waged a high-intensity conflict from orbit, stitching together commercial constellations of small satellites for imagery, communications and targeting.
What this conflict has laid bare, and what a wave of artificial intelligence now amplifies, is that smallsats have become strategic assets on their own with a barrier to fielding them that appears to be falling fast, including for nations that, until recently, had no orbital capability whatsoever.
Explore the August 2026 Issue
Check out more from this issue and find your next story to read.
The case for the small satellite is architectural to begin with. Where the traditional approach concentrated capability into a few pricy platforms, proliferated constellations spread it across dozens, sometimes hundreds, of mass-produced and thus cheaper units. The tactical advantages are considerable: faster revisit times over a given area, no single point of failure, and the ability to absorb the loss of individual satellites without fully compromising the mission. A constellation is, by design, hard to passivate. It is too dispersed to be neutralized by a single strike, and inexpensive enough to be reconstituted on short notice.
It is this logic of disaggregation, as much as cost, that explains why the U.S. Space Force’s Proliferated Warfighter Space Architecture intends to place well over a thousand small satellites in Low-Earth Orbit (LEO) for missile tracking and secure communications. Resilience, once purchased through expensive in-orbit redundancy, is becoming an emergent property of numbers.
Ukraine has served as the live demonstration of this shift. Despite the near-absence of any sovereign space capability, Ukrainian forces have conducted operations underpinned by commercial systems: tens of thousands of Starlink and Starshield terminals sustaining battlefield communications and drone operations, and high-revisit imagery from operators such as Planet, Maxar, Iceye, Capella, and HawkEye 360 delivering all-weather overhead intelligence.
Analysts now describe space as having moved, over the course of this war, from a strategic enabler to a tactical dependency acting as a space-enabled “internet of the battlefield,” without which Ukrainian operations would be severely degraded. Tools fusing drone, satellite and forward-observer intelligence have let troops task imagery, mark targets and coordinate fires in hours rather than days. The lesson has not been lost on other armed forces: the ubiquity of sensing, once the privilege of a few nations, is quickly becoming a tactical commodity.
If proliferation is the structural change, artificial intelligence is its accelerant. The first generation of small satellites largely behaved as “bent pipes,” collecting raw data and relaying it to the ground for processing. Onboard processors can now run AI models that interpret imagery in orbit and downlink actionable insights rather than terabytes of raw pixels. This is a decisive advantage when the link to the ground is congested, contested, or jammed.
The milestones are arriving in quick succession: geospatial foundation models running directly on commercial constellations, satellites carrying multi-customer compute payloads on which several operators run their own models, and the launch of high-end graphics processors to probe the limits of orbital computing. China, for example, has begun deploying its “Three-Body Computing Constellation,” a dedicated in-orbit computing constellation, demonstrating distributed processing and the operation of large AI models in space. For the military user, this collapses the sensor-to-shooter timeline and pushes autonomy to the edge, precisely where decisions must increasingly be made.
Taken together, falling launch costs, mass-produced platforms, and embedded intelligence are doing for military space what mobile networks did for
telecommunications across the developing world: they let latecomers leapfrog. New space nations no longer need a decades-old space legacy to field a sovereign capability: a modest constellation in LEO can now deliver imagery or communications that were the preserve of superpowers a generation ago.
The Gulf states illustrate the model, pairing sovereign-wealth funding with commercial partnerships to build indigenous capability at speed. Europe, chastened by its own dependence, has committed billions to sovereign constellations, from Italy’s IRIDE to the continent-wide IRIS², explicitly framed around national control of defensive and, increasingly, offensive orbital capability. And as a consequence, market analysts now expect defense and civil satellites to overtake commercial deployments, as a new generation of suppliers emerges to serve governments intent on owning their own eyes and ears in orbit.
This democratization, however, is double-edged, and any clear-eyed assessment must hold both sides together. Dependence on commercial providers introduces a vulnerability of a new kind: the operator’s off-switch. The temporary suspension of a shared imagery service to Ukraine, and the reported refusal to extend connectivity over contested territory, both showed how a private actor can hold an effective veto over a battlefield.
Ubiquitous sensing, too, cuts both ways, when high-revisit imagery can be bought by anyone, surprise becomes harder to achieve for attacker and defender alike, and the very transparency that serves one belligerent serves its adversary. And the constellations themselves are now targets: the opening salvo of the war in Ukraine was a cyberattack on a satellite communications network that rippled across European civilian infrastructure, a reminder that proliferation reduces vulnerability without ever abolishing it.
Small satellites, made resilient by numbers and intelligent by AI, are therefore best understood for what they are: a structural rebalancing of who can act from orbit. They will not render the pricey satellite obsolete, and they carry dependencies and vulnerabilities that policy has barely begun to address. Yet, they have already redrawn the map of space power, handing meaningful military capability to new entrants and forcing established powers to think, and to fight, in terms of constellations rather than crown jewels.
The coming years will tell whether the nations now rushing into orbit can convert access into genuine strategic autonomy, or whether, in leaning on the constellations of others, they have merely traded one dependence for another. VS
