Sberbank’s AI-related revenue reached 1.5 trillion roubles, about $18 billion, over the past three years, chief executive German Gref told investors on 12 November, Reuters reported. Russia’s largest bank unveiled its 2027-2029 strategy 29 days after Ukrainian drones struck its data centre, Russia’s largest, on 14 October; Sberbank said the attack caused no major disruptions.
Gref said last year that Sberbank planned to invest 350 billion roubles in AI in 2026, Reuters reported. That is about $4.2 billion, and 70% of the bank’s average yearly AI-related revenue of 500 billion roubles, by our arithmetic.
Business Pill · GPUs
An evergreen Business Engineer explainer on why AI runs on GPUs. It teaches the general idea only and does not discuss Sberbank.
The key insight: As we read the figures, Sberbank is pairing a large AI revenue claim with a capital plan that still puts shareholders first: AI-related revenue of 1.5 trillion roubles over three years equals under 63% of the more than 2.4 trillion roubles it paid in dividends, and its 2026 AI investment plan equals 70% of a year of that revenue, by our arithmetic.
The bank’s Strategy 2029 release, issued in Russian at its investor day, says its strategic goal to 2029 is to become one of the leading architects and builders of Russia’s agentic economy, in our translation. It names commercial use of generative AI as one of the most important directions of the 2027-2029 strategy: besides cutting costs, the bank intends to use AI to create new revenue sources.
The release says technology development will require substantial investment in infrastructure and system reliability. It says Sberbank targets return on equity of no less than 21% a year, keeps 50% of IFRS net profit for dividends, and aims to hold its cost-to-income ratio at 32-35% over the next three years.
Gref said Sberbank plans to accelerate construction of data centres and does not rule out acquiring players in the data-centre market, Finam reported from the investor day.
Sberbank, billion roubles: AI-related revenue over the last three years (1,500, about $18 billion) and AI investment planned for 2026 (350, about $4.2 billion), as stated by chief executive German Gref and reported by Reuters, against dividends paid over the last three years (more than 2,400), per the bank’s Strategy 2029 release. Dollar conversions at Reuters’ rate of 82.8656 roubles; the comparison is our arithmetic.
The Drone Strike and the Revenue Figure
Reuters reported that the 14 October strike appeared to be part of a broader campaign targeting infrastructure linked to Moscow’s AI ambitions. Earlier, Interfax reported that data centres of three Russian companies were hit by drone attacks in October, as we covered on 2 November.
Gref told investors: “The race for leadership in AI is a matter of national sovereignty and our survival”. Reuters reported that he added that Russia needed high-quality AI models and resilient IT infrastructure, and that he did not say how Sberbank would gain access to the latest microchips, a major challenge due to Western sanctions.
The release says Sberbank’s dividend payments over the last three years exceeded 2.4 trillion roubles, as much as the bank paid over the previous 20 years. Set against that, the 1.5 trillion roubles of AI-related revenue equals under 63% of dividends paid, by our arithmetic.
The Structural Read
As we read the release, the bank names infrastructure and system reliability as an investment need in the same strategy that keeps a 50% dividend payout and a 21% return-on-equity floor, and Gref said it plans to accelerate data-centre construction, according to Finam.
As we read the Reuters report, the open variable is hardware rather than demand: Gref did not say how Sberbank would gain access to the latest microchips, which Reuters called a major challenge due to Western sanctions.
German Gref, Sberbank chief executive, as reported by Reuters, 12 November 2026
“The race for leadership in AI is a matter of national sovereignty and our survival”
Three Implications
FOR SBERBANK’S CAPITAL PLAN The release keeps 50% of IFRS net profit for dividends and a return on equity of no less than 21% a year, alongside substantial investment in infrastructure.
FOR RUSSIA’S DATA-CENTRE MARKET Gref said Sberbank does not rule out acquiring data-centre players, Finam reported.
FOR AI HARDWARE Reuters reported that access to the latest microchips is a major challenge for Sberbank due to Western sanctions.
The framework says: “The digital economy rests on a physical foundation that most people in that economy never see.”
As we read it, a drone strike on a data centre and a chip-access gap under sanctions are both versions of that physical foundation. The framework predates Sberbank’s strategy and does not discuss it.
Sberbank’s average yearly AI-related revenue over the last three years, 500 billion roubles by our arithmetic from the 1.5 trillion Gref cited, against the 350 billion roubles of AI investment he said last year the bank planned for 2026, as reported by Reuters. The 70% ratio is our arithmetic.
What Is Not Established
The reports we read do not define AI-related revenue or split it by year; the 500 billion rouble average is our arithmetic. The 350 billion rouble figure is a plan Gref stated last year, as reported by Reuters, not a reported outturn for 2026.
We read the strategy release as republished in Russian by a regional outlet; the bank’s own website did not load for us. Dollar figures use Reuters’ rate of 82.8656 roubles to the dollar. We did not contact Sberbank for this piece.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
Sberbank put its AI-related revenue at 1.5 trillion roubles over three years and set a 2029 goal built on AI agents, a month after the strike on its largest data centre. Its 2026 AI investment plan of 350 billion roubles equals 70% of its average yearly AI revenue, by our arithmetic.
A note on sourcing. The AI revenue figure, the 2026 AI investment plan, the Gref quotation and the account of the 14 October strike come from Reuters’ report of 12 November, read in full via MarketScreener. Strategy targets come from Sberbank’s Strategy 2029 release in Russian, as republished by Obozrevatel.Vrn, in our translation; the data-centre remarks come from Finam. Dollar conversions use Reuters’ rate; averages and ratios are our arithmetic. Sberbank was not contacted for this piece.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
