For a ship working the Northern Sea Route in January, the connection to shore is the difference between a voyage and a gamble. Ice forecasts, engine diagnostics, crew medical advice and the paperwork that clears a port all travel over a link that, in Russian waters, Moscow does not control. A constellation now being assembled is meant to change that, and the coverage it offers will be deliberately narrow. The Financial Times reported on Nov 8 that Russia is pushing ahead with a low Earth orbit broadband network called Zorky, with deployment scheduled to begin in 2026. The project is run through the state space establishment and is positioned as a domestic substitute for Starlink. Its planned satellite count and the area it will cover are both far below the Western system’s, and the first customers are described as remote Russian regions and the Arctic shipping lane.
The modesty is the strategy. A global constellation is a manufacturing problem before it is anything else: thousands of satellites built on a production line, launched on a cadence of weeks rather than seasons, replaced every five years as their orbits decay. Starlink got there by treating satellites as a consumable and rockets as a conveyor belt. A programme that intends to serve one country’s far north does not need to solve that problem, and a programme that cannot solve it should not try. Serving the Arctic is genuinely easier. A satellite in a polar orbit passes over the high latitudes on every revolution, which is why coverage at the top of the world is the cheapest coverage a constellation can buy. A few hundred spacecraft in the right orbital planes can keep a continuous signal over the shipping lane, the mining towns and the military bases strung along the coast, while leaving the rest of the planet to somebody else. That is a defensible engineering choice, and it is also a political one: it accepts a regional role in exchange for a system that can be finished.
Whether it can be finished is the open question. Every constellation is limited by three capacities that have nothing to do with radio: how many satellites a country can build, how often it can launch them, and how many ground terminals it can put in the hands of users. Russia’s launch cadence has contracted over the past decade, and a network that needs regular replenishment is exposed to a single failure in that chain. Analysts who follow the programme describe timetables that have slipped before, and a 2026 start is a start rather than a service.
The ground segment carries the other half of the bill. A satellite is useless without a dish, and a dish that costs more than a year of the service it delivers will not sell in a village with a hundred residents. Starlink solved this by building terminals at consumer-electronics volumes and, in several markets, subsidising them. A state programme funded from a budget under war pressure has less room to subsidise, which points toward a customer list made of shipping companies, resource firms and government agencies rather than households.
Sovereignty, not coverage, is what the system is being bought for. Traffic that runs through a domestic constellation lands at a domestic ground station, where the law of the country applies to it. There is no foreign operator to negotiate with, no terms of service written in another jurisdiction, no risk that a provider will withdraw service during a conflict. Every government that has watched satellite links become a battlefield instrument has drawn the same conclusion, and several have begun paying for the redundancy. That lesson is not hypothetical. The war in Ukraine turned satellite broadband into a piece of military infrastructure, with terminals directing artillery and drones and a provider’s commercial decisions becoming matters of state. The consequence was a rush by governments everywhere to ask what they would do if the link went down. Russia’s answer is a network of its own, and the Northern Sea Route gives it a civilian justification that a purely military programme would not have.
There is an export dimension as well. A satellite internet system built outside the Western supply chain is a product for countries that cannot buy the Western one, and the list of governments in that position is not short. Whether Zorky becomes a commercial offering or stays a domestic utility depends on whether the satellites can be built at a price that allows one, and on whether ground terminals can be exported without the components that sanctions restrict.
For Starlink, none of this changes the global map. A regional constellation serving the Russian Arctic does not compete for a customer in Ohio or Osaka, and the Western system’s advantage in launch cost and manufacturing volume is not threatened by a programme an order of magnitude smaller. The market being divided is not the consumer market. It is the market for sovereign connectivity, where the buyer is a state and the product is the assurance that nobody else holds the switch. That market is smaller than the one Starlink built, and the margins inside it are set by politics rather than by price.
What to watch in 2026 is narrow and specific. The first satellites will tell observers how fast the programme can actually build, and the first terminals will tell them who the customer really is. A network that connects a few thousand Arctic users is a modest thing. A network that lets a government say it no longer depends on anyone for the link to its own territory is something else, and the difference will not show up in the coverage map.
