Russian firms plan more cloud and AI spending

The shift matters economically because it points to a structural response to constraints in Russia’s technology supply chain. According to the poll of 503 respondents, 71% of organizations face at least one problem when buying or operating their own equipment. The biggest barrier is cost: 28% said IT hardware is too expensive, especially RAM and graphics processors, the components most closely tied to modern AI workloads. That is a clear indication that the AI buildout in Russia is becoming less about owning chips and servers and more about accessing compute as a service.

For providers of cloud infrastructure, this creates a demand tailwind even in a constrained economy. MTS Web Services, through its MWS Cloud unit, is effectively pointing to a market where companies are trying to preserve digital capability without large capital outlays or procurement headaches. That is favorable for cloud vendors, data-center operators and domestic software platforms that can bundle AI tools with storage and database services. It is less favorable for enterprises that wanted to modernize in-house on their own timetable, because dependency on third-party infrastructure can raise recurring costs and limit flexibility.

For investors, the story is less about a single survey reading than the business model it reinforces. AI demand in Russia appears to be shifting toward subscription and usage-based services, a setup that can support recurring revenue growth for local cloud providers if they can secure enough hardware and power. But the opportunity is bounded by supply constraints, currency pressure and the broader sanctions environment, which still make advanced accelerators and memory difficult and expensive to source. That makes the upside real but uneven: demand may be there, but supply and pricing could determine who captures it.

The broader narrative is that Russian companies are adapting to a capital-scarce, supply-constrained technology environment by outsourcing more of the AI stack. If that pattern persists, the near-term winners will be cloud and service providers; the losers will be firms trying to build large in-house AI capacity, and any vendors dependent on imported hardware.