Starlink in Ukraine: What’s Working, What Just Broke, Who Pays

Late on Tuesday, October 13, Ukrainian military units began reporting trouble with their satellite links. By Wednesday morning, Serhii “Flash” Beskrestnov, a defense-technology adviser to President Zelensky, said publicly that speed restrictions had been placed on terminals while they were in motion, without any notice to Ukraine’s forces. The practical effect was severe. Interceptor drones, the cheap fast aircraft Ukraine uses against Russia’s Shahed-type attack drones, rely on satellite internet, and a terminal on a fast-moving airframe that drops its connection is a drone that no longer reaches its target.

The timing made everything worse. According to Axios, the restrictions arrived hours after Ukraine’s military failed to receive an early warning about a Russian ballistic-missile attack, an alert that Ukrainian officials say the U.S. has supplied for years. Kyiv was hit by missiles on a second consecutive night. Al Jazeera, citing Reuters and AFP, reported at least nine dead and damage to electricity and heating infrastructure.

When we lined up the Axios, Al Jazeera, Kyiv Post and Kyiv Independent accounts side by side, we found that the speed figures don’t agree. Axios reported that the connection dropped when a drone passed 160 km/h and, in some cases, when it passed 80 km/h. Kyiv Post, citing Defense Ministry sources, said units lost connections at different speeds ranging from about 72 to 161 km/h. Al Jazeera described a limit for terminals traveling faster than 100 mph. Those are all roughly the same threshold viewed through different units and different anecdotes, but the spread is real: a fixed 160 km/h cap is not the same experience as a link that fails at 80 km/h.

That spread explains a line from a Ukrainian official quoted by Axios, who said conditions were better than that morning but that 160 km/h “is not enough” to intercept some Russian jet-powered drones. The Defense Ministry also told journalists that actual speeds available during the day were lower than the nominal cap. If you are trying to judge how bad this was, the honest summary is: bad enough to disrupt several missions, partially eased by Wednesday evening, and reported as restored on Thursday morning with unanswered questions.

Here is the part most explainers skip. Musk’s framing, that Starlink is a commercial product restricted to defensive use while Starshield belongs to Washington, is internally consistent, but it also changes who is accountable. If the restriction was on Starshield, the decision sits with the U.S. government. If it was on Starlink, it sits with a private company that Ukraine has leaned on since 2022. Ukraine says both were hit, which is why Fedorov, who ran the Defense Ministry through the whitelist rollout, said the focus must now be on removing the Starshield speed limit entirely.

There is also an older thread. In the 2025 dispute, Musk wrote that no matter how much he disagreed with Ukraine policy, Starlink would never turn off its terminals and would never be used as a bargaining chip. This week’s incident did not turn terminals off, and nothing published shows it was a bargaining move. But it demonstrated that “never turn off” and “never degrade” are different promises, and a government that depends on a service for interceptor drones cares about the second one as much as the first.

At the end of January, Ukrainian officials found Starlink terminals mounted on Russian long-range attack drones. This mattered because Starlink-linked drones fly low, are hard to shoot down, and are not disrupted by the GPS and radio jamming that Ukraine uses against older drone types. Defense officials said Russian operators were bypassing Ukraine’s electronic defenses and striking deeper. Ukraine’s Defense Ministry contacted SpaceX, and Musk replied within hours that service was active and more terminals were on the way, and then agreed to cooperate on verification.

The solution Ukraine’s Cabinet adopted was a whitelist: by default no terminal works, and only terminals that have been verified and submitted function. Fedorov announced the system on February 2, registration opened on February 3, and on February 5 he said unapproved Russian systems had been blocked. Reuters reported a Russian military blog describing a large failure of Russian-side Starlink terminals starting Wednesday evening, and a Ukrainian military source said Russian units were scrambling to use domestic satellite terminals instead.

In our review of the Ministry of Defence’s published instructions and its follow-up FAQ, the pattern we found was that most frustration comes from three things: the daily update cycle, mismatched identifiers, and ownership changes. Fedorov himself told users in February that a terminal already submitted but not yet working should simply wait, because the whitelist was refreshed once per day at that point. The Ministry also said it was working toward real-time submission. We could not find a published confirmation that real-time processing has been completed, so treat the once-a-day statement as the last confirmed state.

Identifier mismatches are the second trap. The Ministry’s guidance asks for the kit number from the box label, the terminal’s unique ID, and the Starlink account number. A buyer who purchased a secondhand dish without its box, or who inherited a terminal under someone else’s account, can run into trouble. And the third trap is what we flagged in the table: registration is tied to a person or entity, not to the hardware forever. If you sell or give a dish away inside Ukraine, the new owner has to go through the process on their own.

One more caution, because it affects how much weight to give some of the louder headlines. A report that circulated on aggregator sites this week, citing the Daily Telegraph and frontline commanders, claimed that authentication failures under the whitelist have left some platoons without links and degraded medium-range drone operations. We could not independently confirm the details, and the Defense Ministry’s own public explanation of the October restrictions points to speed limits rather than whitelist failures. It is plausible that both problems exist. It is not established that they are the same problem.

Start with the history so the present makes sense. Ukraine’s then digital minister, Mykhailo Fedorov, asked Musk for terminals on February 26, 2022, and the first shipment arrived two days later. USAID later said it delivered 5,000 terminals through a public-private arrangement with SpaceX, of which, according to a report at the time, 3,667 were donated by the company and 1,333 were bought by USAID. In late 2022, SpaceX told the Pentagon that roughly 85 percent of an early batch of about 20,000 terminals were fully or partially funded by outside sources, including the U.S., U.K. and Polish governments. SpaceX also said at that time that it was covering the costs of a large share of the service.

Poland became the anchor funder. In 2024 Polish officials said Warsaw funded subscriptions for more than 20,000 terminals. In February, Fedorov wrote that Poland continues to fund more than 29,000 terminals, and the Ukrainian Ministry of Digital Transformation described the figure as the total Poland has delivered since the full-scale invasion and continues to support financially. Poland’s foreign minister, Radoslaw Sikorski, wrote in 2025 that the cost is about $50 million a year. A Polish digital-affairs official also noted that Poland technically loans the terminals to Ukraine to avoid tax complications, which is a detail worth knowing if you ever wonder who legally owns a given dish.

We found that the biggest source of confusion online is mixing these numbers up, so it is worth being explicit about which one answers which question. “More than 50,000 terminals” is the reported count of terminals operating in Ukraine, a figure repeated by Euronews, the Recorded Future News team and others around the February whitelist announcement. “More than 29,000” is the number Poland says it funds. They overlap rather than add. And there is no public breakdown of the remainder by funder that we could find; the earlier crowdfunding and NGO figure of around 6,000 comes from a single analytics site and should be treated as an estimate.

The cost side is similarly layered. In 2022 SpaceX described different service tiers worth very different amounts per month, with most terminals contracted at the lower tier. Those figures are three-plus years old and should not be used as current prices. The only current cost anchors we found are Poland’s roughly $50 million annual figure for its share and a charity’s listed price of about 18,000 hryvnia, roughly $435, for one terminal bought in Ukraine.

The intuitive move is to buy a dish and mail it. Resist it. A terminal registered to a U.S. account is a problem for whoever receives it, because the Ukrainian whitelist asks for identifiers tied to an account, and Ukraine’s own guidance says a changed owner has to go through verification again. International shipping is slow, customs can intervene, and a package that arrives after your assumptions have gone stale helps no one. We could not find an official Ukrainian channel designed to receive donated U.S. terminals, so if someone offers you one, ask hard questions.

Money to an established fund is simpler. Charities already buy terminals in-country, which avoids shipping and puts the device straight into the local verification process. Two examples from our research: the Serhiy Prytula Charity Foundation reported providing 27,702 communications devices, including Starlink terminals, repeaters and radios, and it has a U.S. arm; and Come Back Alive reported directing more than 18.6 million hryvnia in donations toward communication gear, including Starlink terminals, for a security-service unit. UNITED24, the government-run fundraising platform, has also run campaigns that included terminals.

Your home dish and the terminals in Ukraine share hardware heritage but not service terms. Nothing we found links the Ukraine restrictions to residential coverage in the United States. What did change for you is price: SpaceX raised U.S. Residential and most Roam plans, with new subscribers affected immediately and existing customers seeing the higher rates on or after June 18. Starlink also revised its plan names earlier in the year, so older guides may mislabel them. Residential Lite became the 200 Mbps plan, and the flagship became Residential Max.

Prices below reflect the figures reported after that increase. Starlink runs frequent promotions that vary by address, so treat these as list prices and verify on the official site before ordering.

On November 24, Kyivstar announced that Starlink Direct to Cell was available for testing to all of its subscribers, making Ukraine the first country in Europe to offer it. The service lets ordinary phones connect to satellites when ground coverage is down. In its first phase it carries text messages, with light data, voice and video planned for later. Kyivstar said it was initially compatible only with Android phones that support 4G LTE, with iPhone support expected, and that it comes at no extra cost to subscribers. The network also needs a clear view of the sky, so it does not work indoors.

Why it matters for the bigger Starlink story: Kyivstar serves about 22.5 million customers and already has batteries and generators designed to keep towers running for up to ten hours without grid power. Direct to Cell does not replace a terrestrial network and was never meant to. What it offers is a last-resort message channel when towers go dark during attacks on the grid, which is the scenario Ukraine has lived through repeatedly. If you are a U.S. reader with relatives there on Kyivstar, this is the feature worth asking them about.

The comparison that does the most work is satellite counts. In 2025 reporting, Starlink was described as operating roughly 7,000 satellites against about 600 for Eutelsat. That is not a gap that a single funding round closes. When the Eutelsat CEO said a dependence like this “can be decided in the White House or Mar-a-Lago,” the quote was a sales pitch and a warning at the same time. And there is an irony in the UASAT story: its first satellite was listed as launching aboard a SpaceX Falcon 9 rideshare mission. Even Ukraine’s effort at independence began on the incumbent’s rocket.

The realistic near-term path is diversification rather than replacement: more OneWeb and other terminals for the highest-priority links, Direct to Cell for civilian messaging resilience, and a long build-out for sovereign constellations. If you are trying to read the news, watch for announcements about terminal orders and funding commitments from European governments, not for claims that Starlink is about to be dropped.