Starlink and Why Russia’s War Just Slowed Overnight

The painful lesson of fighting on infrastructure you don’t own

My personal photo of a Starlink flight over my house

There’s a moment early in this war most people forgot about. A strange little scene in September 2022, when an uncrewed Ukrainian sea drone washed ashore in occupied Crimea like a message in a bottle.

At the time, analysts scrambled to explain it: A navigation error? A battery failure? A classified system glitch no one would ever publicly describe?

Now we know the truth.

That drone didn’t fail. It was cut off.

AFU

According to excerpts from Walter Isaacson’s biography of Elon Musk, the head of SpaceX ordered his engineers to disable Starlink coverage around Crimea, right as Ukraine launched the largest maritime drone strike of the war against the Russian Black Sea Fleet.

Ukrainian operators lost connectivity, the drones drifted, and one of them ended up on camera for the whole world to puzzle over.

The justification Musk offered Isaacson was that he feared nuclear escalation.

The balls on this guy… A civilian tech company CEO interfering with Ukrainian military operations because he, personally, feared Russia would nuke us if Ukraine won too much?

What the actual hell, man.

As if Russia would start throwing around strategic nukes because Ukraine attacked a military target, on Ukrainian territory, recognized as such by the United States, European Union, and almost every country outside Russia’s propaganda ecosystem.

And remember, this was the old Elon Musk… The one beloved by millions of people as the savior of humanity. He would solve the climate crisis with his electric cars and get humans to Mars. So, suffice it to say I was extremely unpopular when I attacked Musk in an article on Medium at the time.

Some notable science writers sent me messages saying, “How dare you attack Elon!”

Of course, we now know Elon as the guy who paid to get Trump elected then single-handedly took a torch to millions of dedicated career civil servants through the ridiculous department of government efficiency.

Still, when Musk hit the kill switch and turned off Ukrainian access to Starlink in 2022, it was an early sign of the bigger structural problem: a modern war leaning heavily on a civilian satellite network owned by a single private American company. A company whose CEO sometimes behaves like he personally adjudicates world affairs from the hinge of his phone screen.

But the Starlink-Crimea incident isn’t the story here. It’s the prelude.

The real story starts when Starlink stops working for Russia. And everything Moscow built on top of it collapses in real time.

The quiet backbone of a modern war

From 2022 through early 2024, Starlink was infrastructure. The nervous system of Ukraine’s war effort. A platoon in a trench could push live drone video to brigade headquarters. Logistics teams in shattered towns could track ammunition flows. Mortar teams could receive targeting updates with satellite latency measured in tens of milliseconds. Command posts no longer had to choose between staying hidden and staying connected.

Starlink did three things no previous system could do simultaneously:

It was high-bandwidth.

It was mobile.

It was resilient.

Ukrainian units used it everywhere. Cities. Forests. Muddy fields. Basement command posts. Moving vehicles.

It worked because SpaceX had already solved the engineering nightmares that legacy military systems spend decades wrestling with. A massive constellation in low Earth orbit, constantly replenished, flooding the battlespace with reliable coverage.

And Russia used it too.

By late 2023, Russian units were operating Starlink at scale. Some terminals were captured on the battlefield. Others were bought through smugglers in the Caucasus or Central Asia. Some were reprogrammed by Russian engineers.

And a Russian drone unit went a step further and mounted Starlink terminals directly onto long-range fixed-wing UAVs. It let them bypass Ukrainian jamming, fly deeper, and maintain control links Ukraine couldn’t interrupt without shutting down its own forces.

Russia never said this publicly. They pretended their glorious sovereign military didn’t need Western civilian tech. But their soldiers knew the truth. Their milbloggers knew. The Ukrainian General Staff definitely knew.

The entire Russian tactical drone ecosystem leaned on Starlink like a crutch.

And in early February, SpaceX kicked the cane out from under them.

The day the lights went out

Last week, SpaceX pushed a quiet but decisive shift: only verified terminals would remain online.

Everything else… grey-market Starlinks, smuggled terminals, captured Ukrainian units, Russian field improvisations, went dark. Movement-based restrictions cut off terminals used on fast-moving platforms, especially drones.

It happened fast. Russian channels lit up like a Christmas tree.

Assault units complained they were suddenly blind. Drone feeds froze mid-flight. Artillery spotting broke down. Fixed-wing drones lost control links. FPV coordination disintegrated into frustrated joystick shaking.

One Russian milblogger said it outright: “Starlink was the easy way.” Not even in a defeatist tone. In a tone of exhausted honesty.

And the uncomfortable truth underneath that complaint is simple: Russia never built a replacement.

Ukraine treated Starlink like a temporary fix and spent two years building its own military digital backbone around systems like DELTA. Russia treated Starlink as a permanent shortcut and built doctrine around it.

When the shortcut vanished, the doctrine cracked.

There’s a reason Russia is uniquely vulnerable to a communications cutoff. Its entire command philosophy depends on centralized control.

Orders flow downward. Approval flows upward. Fire requests move through layers of confirmation. Drone operators are tied to higher headquarters for targeting authority. Nothing moves easily without a digital tether.

That system is brittle even on a good day. When the comms go out, it seizes up.

Ukraine’s approach is the opposite. Mission command. Distributed decision-making. Junior leaders expected to improvise. A platoon can fight independently for short periods because the doctrine is built for decentralization.

Take away comms from a Ukrainian unit and they degrade. Take away comms from a Russian unit and they freeze. That’s a big difference.

Interceptions over the past week confirm this: Russian assault squads near Kupiansk stalled, units near Vovchansk attacked without drone support and retreated in confusion, and FPV crews tried to improvise visual spotting, which is like playing darts with a blindfold. One partisan group, Atesh, claimed a Russian assault element got wiped out by friendly fire after losing communications entirely. That report isn’t verified, but it fits the pattern: units without awareness become hazards to themselves.

The drone war suffered the most immediate damage.

Russian FPV teams built their tactics around real-time video. Their long-range UAVs relied on Starlink links to fly well beyond the reach of their own EW-vulnerable datalinks. Their artillery spotting depended on steady, high-throughput uplinks.

Take away the satellite backbone and the entire machine sputters.

Russia tried to use cellular networks as a backup. Those networks collapsed under Ukrainian EW pressure years ago.

Radios? Jammable. Fiber? Requires trenching under fire. Wi-Fi bridges? Line-of-sight only and vulnerable to weather. Fiber-optic drones? Doable, but they have their own limitations.

Satellite internet was the hormonal system of the Russian tactical battlefield.

And now the hormones are gone.

The predictable line from Russian talking heads was that Moscow could simply switch to its own satellite operator, Gazprom Space Systems.

On paper that sounds reassuring, right up until you look at what Gazprom actually runs.

Its Yamal constellation is a handful of mid-size geostationary communications satellites parked 36,000 kilometers above Earth at fixed longitudes like 49° East. The newest, Yamal-601, carries 70 transponders in C- and Ka-band and throws 32 Ka-band spot beams over western Russia and western Siberia to feed corporate networks, regional administrations, and fixed VSAT customers through a small number of ground hubs.

It is a perfectly respectable GEO comsat for broadcast and business broadband; it was never designed to support thousands of small, moving terminals bolted to pickup trucks, command posts, and drones on a live front line.

Starlink is built on the opposite philosophy.

Instead of six geostationary platforms covering continents from 36,000 kilometers away, it throws more than nine thousand small satellites into low Earth orbit around 550 kilometers up, forming a dense mesh that hands off traffic between satellites and down to gateways with latencies around 25 to 35 milliseconds, roughly in the same range as decent terrestrial broadband.

Traditional GEO systems like Yamal live in the 500 to 700 millisecond range, which is fine for TV and bulk data but disastrous for real-time remote piloting, low-altitude drone flying, or live fire coordination.

On top of that, Gazprom’s fleet serves a domestic and regional market with limited beam capacity, while Starlink was built from day one to handle millions of concurrent users scattered across the planet.

Russia is trying to replace a global LEO broadband mesh with a few GEO TV-plus-internet birds and a couple of teleports.

It is not a lateral move.

China gets mentioned as a theoretical backup. The odds of Beijing casually handing over high-bandwidth resilient satellite internet with political, military, and intelligence implications are low. Integration would take months at best. The war moves daily.

So, this isn’t a problem solved by buying gear. It’s structural. And I believe it’s a turning point.

Russia is learning, very slowly, what Ukraine already learned in 2022. You can’t conjure battlefield connectivity out of thin air.

Does this mean we should praise Elon Musk for finally doing something he should have done years ago? No.

Not long after the Starlink restrictions were observed on the Russian frontline, Russian state host Vladimir Solovyov appeared on television suggesting that Russia could detonate a nuclear device in orbit to take out the entire Starlink constellation.

It was theater. It was also revealing.

Countries don’t threaten space-nuclear detonations unless they’re out of ideas. This was frustration disguised as bravado.

For a nation that constantly brags about sovereignty, self-reliance, and “import substitution,” admitting that its frontline units depend on an American civilian satellite service is embarrassing.

Moscow would rather fantasize about orbital weapons of mass destruction than acknowledge that Ukrainian bureaucrats in Kyiv now determine which Starlink terminals stay online and which ones don’t.

The anger makes sense once you understand the scale of Russia’s dependency.

Starlink wasn’t a convenience. It was a prosthetic.

Remove it and Russia goes back to stumbling.

Licensed by the author

Kyiv’s side of the switch

Ukraine capitalized on Musk’s crackdown of unauthorized terminals. On February 2, Ukraine’s Cabinet of Ministers approved a resolution tightening Starlink access across the country.

Author’s note / clarification:
SpaceX began enforcing technical restrictions on Starlink usage first, including speed and unauthorized-terminal limits. Ukraine then formalized those restrictions through a government verification and whitelist system, turning them into state-controlled battlefield infrastructure.

The policy sounds boring, but its impact is enormous. Only verified terminals stay online. This includes military units, government agencies, civilian rescue services, registered businesses, and individual citizens who complete a quick identity-linked registration.

Ukraine is doing it to purge any Russian residual usage.

Civilian registration is painless.

A Diia login or one quick visit to an Administrative Services Center. It feels like renewing a driver’s license.

No Soviet paperwork. No pointless queues.

The military process is separate.

Terminals are registered through secure channels inside DELTA, the same wartime digital backbone that tracks Russian troop movements and artillery patterns. Units don’t surrender personal logins or hand over terminals. They simply confirm IDs so the network knows who’s friendly and who’s not.

Some Ukrainian units did lose access temporarily. Some had to scramble to register older terminals. Some faced delays. But Ukraine expected this.

The Defense Ministry prepared workflows months in advance, which shows remarkable foresight in avoiding friendly battlefield disruptions.

The slight disruption that did occur wasn’t catastrophic, but the gain in battlefield asymmetry was enormous.

The drone war without Starlink

This is where the consequences hit Moscow hardest.

Over the last year, Russia pushed drone-infantry integration aggressively. FPV units became the tip of the spear for close attacks. Larger fixed-wing drones conducted deep reconnaissance and long-range strikes.

Drone operators piped targeting data straight into artillery formations.

All of that required reliable connectivity.

Starlink gave Russian drone teams something their own military systems couldn’t provide: bandwidth at range. When Russian EW units jammed their own troops’ drones accidentally, a common occurrence, the fallback became Starlink.

When Ukrainian EW suppressed Russian C2 channels, Starlink kept their drones alive.

That whole safety net just disappeared.

Ukrainian reporting from the Joint Forces Group indicates exactly what you’d expect: Russian small assault teams are now advancing without FPV support.

Let that sit in your brain for a moment… Right now, Russian assault units are attacking without FPV support!

Fixed-wing drones lose command links during maneuver.

Long-range drone strikes are less precise.

Coordination between infantry and UAV teams is unreliable.

Weather makes everything worse because, without satellite connectivity, clouds and snow don’t just hide terrain. They sever digital lines.

This is a systemic breakdown for the Russian military in Ukraine.

Russian drone warfare grew dependent on a commercial system they didn’t own and couldn’t control. Ukraine forced that system to identify its users. Russia’s presence on that system evaporated.

Russia’s “alternatives” are confessions

When Russian analysts began publicly listing alternatives to Starlink, the exercise sounded less like problem-solving and more like reluctant self-indictment.

The options floated were stark: lay fiber-optic cable under artillery fire, build ad-hoc Wi-Fi relays between bombed-out buildings, fall back on older digital radios with bandwidth measured in kilobits, or launch satellite constellations that do not yet exist.

These were described as pathways forward, but none were presented with confidence, and none addressed the immediate reality of a frontline that had come to depend on resilient, mobile, high-bandwidth connectivity.

That absence of confidence mattered.

Even prominent Russian pro-war voices acknowledged it. Colonelcassad stated openly that there are no substitutes available today that operate at Starlink’s level.

That assessment wasn’t hyperbole. Starlink is a solved engineering problem. Its alternatives are not. Russia may eventually put more satellites in orbit, trench fiber in limited sectors, improve radio discipline, or kludge together local relay networks.

All of that shit is technically possible. None of it is fast, and none of it is cheap.

This war has punished slow adapters since its first week, and Russia built much of its recent tactical effectiveness around a shortcut it didn’t control.

Now it has to operate at its natural Russian pace again, and that pace is slow.

The strategic picture

Okay, so, this isn’t just about Elon Musk, and focusing on him obscures the real lesson.

Starlink is a symptom of a much larger shift in modern warfare: the growing dependence on civilian digital infrastructure for basic military survival. Satellite internet, commercial drones, cloud services, encrypted messaging platforms, and private-sector imagery were never designed for combat. They became instruments of war because states reached for what already existed and worked at scale.

This is exactly what happens when you let the defense procurement system get wildly out of control. What did you think was going to happen when you allow Boeing to charge the government $250,000 for a toilet seat in a C-5 Galaxy?

Opportunists like Musk swoop in with a much cheaper consumer version that the government ultimately ends up using anyway.

Ukraine treated these systems as tools and built redundancy, doctrine, and adaptation around them.

Russia treated them as substitutes for military infrastructure it never fully developed. That distinction matters. Starlink’s restrictions highlight a basic risk of modern conflict: if you rely on a system you do not control, you are borrowing capability rather than owning it.

When the owner changes the rules, the bargain disappears.

Russia will adapt eventually. It always does. Engineers will be reassigned. Stopgap systems will be fielded. Workarounds will appear. Improvisation is not a uniquely Ukrainian trait.

But the damage is already visible across the battlefield. Russian units are operating at a slower tempo, with worse coordination, reduced precision, more frequent friendly-fire incidents, less initiative at lower levels, and a growing number of stalled assaults that never quite cohere into momentum.

Wars are rarely decided by a single switch being flipped, but sometimes a switch exposes which side was living on borrowed capability.

Russia was. Ukraine wasn’t.

When the rules changed, one side cursed the darkness. The other adjusted and kept moving.

Shortcuts in war are convenient, right up until someone takes them away.

Слава Україні!