Nearly Half of Russians Solve Tasks via AI Answers in Search
The HSE School of Communications and OMI found out: nearly half of Russians read AI answers directly in search results, and Alisa AI by Yandex leads among neural networks with a 40% regular audience.
Almost every second internet user in Russia solves their tasks by reading an AI answer directly in search results — this is the conclusion reached by the HSE School of Communications and the company Online Market Intelligence (OMI) based on a study of user search experience.
Artificial intelligence and classic search no longer exist separately; they have become parts of a single user journey. Users in Russia already perceive AI answers in search results as a natural part of search and use them alongside the familiar list of links.
Search remains one of the most popular entry points to the internet: 74% of users begin solving everyday tasks on a search engine page. But the familiar "query — list of links" model is no longer sufficient. Over the past year, the scenario of reading an AI answer in search results has taken hold: 34% of users have started using it more often, and 27% regularly ask the AI follow-up questions. Only 26% of respondents noticed no changes in their behavior.
Tools divide up tasks
Classic search and AI are not competing but are distributed across different types of tasks. Classic search results hold their position where a specific object needs to be found: when choosing a product or service, 33% of respondents use them. When searching for an exact fact, AI answers have nearly caught up with classic search — 30% versus 31%.
Where information needs not only to be found but also processed, AI's advantage becomes more apparent. For answering a question with multiple possible scenarios and analyzing documents, 27% of users choose AI; for exploring a complex topic — 26%. With a creative or unconventional task, 23% go straight to a separate AI chat. If a task requires deep exploration, 18% of study participants open a separate chat in advance.
"AI is not becoming a replacement for search but is logically integrating into the user journey. The user starts with a query, receives a compiled answer, and, if necessary, clarifies the details. Search is evolving not from links to AI, but from finding information to solving tasks," notes Tatyana Sokolova, associate professor at the HSE School of Communications.
Alisa AI — market leader
In the market of standalone neural networks (separate chatbots), Alisa AI by Yandex is the clear leader. 97% of respondents know it — more than any other neural network. 82% of those surveyed have used Alisa at least once, and 40% turn to it regularly. Users most often trust Alisa with searching for information and explaining complex topics.
Sber's "GigaChat" overtook ChatGPT in terms of recognition: 84% and 86% respectively — however, judging by the wording of the data, recognition of both neural networks is close. At the same time, despite restrictions and disruptions in the operation of foreign services in Russia, ChatGPT still leads "GigaChat" in usage experience: 54% versus 49%. In terms of regular audience share, the figures are 16% and 12% respectively.
Usage scenarios
The main scenario for working with AI remains searching for and gathering information — 48% of study participants named it. Respondents are ready to entrust artificial intelligence with more complex tasks as well: 29% turn to AI to explain complex topics, 27% expect ready-made conclusions from it, and another 19% use it to compare different solution options. From 12 to 14% of users ask AI to help formulate a task, check a result, or prepare a draft.
Young people adapt to the new search format faster than anyone. Among users aged 18–24, 53% have started reading ready-made AI answers in search results more often, and 37% immediately switch to a separate chat to solve complex tasks — a figure twice the average across all users. The higher a user's technology skills, the more actively they combine tools, choosing the right one at each stage.
Methodology
The study covered men and women aged 18–65 — residents of Russian cities who use the internet. A U&A (Usage & Attitude) methodology was used with online data collection (CAWI) via OMI panels. The field stage took place in July–August 2026: 1000 respondents participated in the quantitative survey, and another 23 in the qualitative one; the average interview took 15 minutes.
More from Алексей Черных
159 articles
Related articles
- Pinduoduo and Anthropic paid Kazakhstan 8 billion tenge in taxes
- Disney Will Show Super Bowl 2027 Through the Eyes of Star Wars Characters
- Omnicom: Creative Falls for Third Quarter, Media Grows 10%
- 51% of teens call brand slang flirting
- Elvis Presley Song Rights Holders Demand $41.1 Million from Anthropic