Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
ReleaseInsight · On RadarRB.RU→ original

AI Could Add Up to 2.6 Trillion Rubles a Year to Russian Companies, Not 680 Billion

Russian businesses currently gain about 680 billion rubles a year from AI, and Data Insight and SberMarketing expect that figure to grow over one to three years. In the base scenario, where AI is connected to CRM and customer service, the impact reaches 1.8 trillion rubles. In the optimistic scenario, where agents accept requests and launch the next stage on their own, it reaches 2.6 trillion rubles, 3.8 times the current level. That is why companies need to decide which actions they are willing to hand over to agents.

AI-processed from RB.RU; edited by Hamidun News
AI Could Add Up to 2.6 Trillion Rubles a Year to Russian Companies, Not 680 Billion
Source: RB.RU. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceRB.RU ↗
Editor-in-Chief
Key metric #1The estimate covers industries that account for about 85% of Russia's gross value added
Key metric #2Automating payments and contract changes is harder than automating routing
Key metric #3X5 is rolling out chains of AI agents in marketing and building a system to manage them

Discussions about how much money AI brings usually rely on polished slides. Here, the numbers at least break down by scenario, and they show what the result depends on. Analysts at Data Insight and SberMarketing estimated the impact of artificial intelligence on Russian companies: in 2026 it amounts to about 680 billion rubles a year, and over the next one to three years it could grow 3.8 times, to 2.6 trillion rubles. Kommersant wrote about the study.

What was measured and where the money comes from

The total includes cost reductions and productivity gains. Companies spend less time and fewer resources on routine work and manage to process more tasks. The study covers trade, the financial sector, telecom and digital services, industry, real estate and market services. In 2024, these industries produced about 85% of the country's gross value added.

The authors looked at processes with repetitive operations and measurable indicators. There are three areas:

  • sales and customer service: handling requests, routing inquiries, recommendations and assisting employees;
  • personnel management: mass recruitment, internal services, training;
  • general management: preparing analytics and passing information along for decisions.

Data Insight partner Fyodor Virin explained that the effect can be measured by concrete results. Employees process customer requests faster, spend less time on paperwork, and routine operations cost less.

Two scenarios: assistant or executor

The base scenario yields 1.8 trillion rubles a year. In it, AI is built into work processes: it gets access to company data and is connected to CRM (the system that stores customers, communication history and deals), customer service and marketing platforms. The model suggests, while the decision and the action remain with the person.

In the optimistic scenario, worth 2.6 trillion rubles, agents not only advise but also carry out certain steps themselves. An example from the study: an agent accepts a request, determines whom to pass it to, and launches the next stage without an employee. People do not need to manually transfer data between systems or forward tasks. To achieve this effect, companies will have to rethink how responsibilities are divided between humans and AI, and assess how many tasks an employee can complete in the same amount of time.

The gap between the scenarios is almost 800 billion rubles. It does not come from a smarter model, but from a company's willingness to change its processes and give agents more autonomy.

Limitations and what to do in practice

The Alliance for Artificial Intelligence noted that an agent's autonomy does not mean any operation can be handed to it. Distributing requests between departments is simpler and safer than processing a payment or changing contract terms. Therefore, companies should limit in advance the list of actions available to the agent and verify the results of its work. High-quality data and integration with internal systems are also needed.

Yandex AI Studio head Artur Samigullin pointed to another side of the issue. The effect depends on what the company spends the freed-up time on: serving more customers, releasing products faster, or shortening request queues. If the time simply dissolves, no savings will show up in the report. T1's AI director Sergey Golitsyn named readiness to change work processes as the main condition for large-scale adoption.

X5 is an example of movement in this direction. Development director for AI Mikhail Neverov told Russian Business that the company is rolling out chains of AI agents in marketing and developing its own system to manage them. The difference from individual assistants is simple. When an employee works with several assistants, they pass tasks from one to another themselves, check the answers, and transfer data. In an agent system, AI carries out the chain of actions on its own, while the person sets the task, defines the rules, and determines what is permitted without approval.

For developers and product managers, the takeaway is this. It is better to start with processes that have repeatable operations and a clear metric, for example request routing. The agent's permissions need to be spelled out explicitly: what it does on its own and what requires confirmation. The forecast remains the analysts' estimate, not a fact, and it covers one to three years. But it shows where to look for money: in restructuring processes, not in swapping one model for another.

Story

Key Questions Answered

What was measured and where the money comes from?

The total includes cost reductions and productivity gains. Companies spend less time and fewer resources on routine work and manage to process more tasks. The study covers trade, the financial sector, telecom and digital services, industry, real estate and market services. In 2024, these industries produced about 85% of the country's gross value added. The authors looked at processes with repetitive operations and measurable indicators.

What are the key details on Two scenarios: assistant or executor?

The base scenario yields 1.8 trillion rubles a year. In it, AI is built into work processes: it gets access to company data and is connected to CRM (the system that stores customers, communication history and deals), customer service and marketing platforms. The model suggests, while the decision and the action remain with the person. In the optimistic scenario, worth 2.6 trillion rubles, agents not only advise but also carry out certain steps themselves.

What are the key details on Limitations and what to do in practice?

The Alliance for Artificial Intelligence noted that an agent's autonomy does not mean any operation can be handed to it. Distributing requests between departments is simpler and safer than processing a payment or changing contract terms. Therefore, companies should limit in advance the list of actions available to the agent and verify the results of its work. High-quality data and integration with internal systems are also needed. Yandex AI Studio head Artur Samigullin pointed to another side of the issue.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Private AI Builders Community · Vibe Club

Ship AI Faster with Founders, Engineers & Weekly Architecture Reviews

Inside Vibe Club (hamidun.com/club) — weekly live masterminds with Zhemal Khamidun, Tech Radar, full LMS access, and 370+ production Claude Code skills.

Join Vibe Club →Explore Hamidun Academy
What do you think?
Loading comments…