- Статьи
- Internet and technology
- Let's go our own way: Russian search engines will not abandon classic links following Google
Let's go our own way: Russian search engines will not abandon classic links following Google
Russian search engines do not plan to completely replace human-written content with comments from AI agents. Google has announced a major transformation of its AI Mode algorithm and other AI infrastructure products as part of a major update. Some of the changes will take effect on May 26. The innovation reformats the paradigm of using search engines, as it puts AI agents in the first place and practically means abandoning copyrighted content. In Russia, they also plan to increase the role of neural networks in web surfing, but they are not ready to exclude user materials and classic links from responses to queries. What will happen to the Internet after Google's innovations and how domestic companies react to this is in the Izvestia article.
How Internet search will change
Large Russian Internet companies do not plan to abandon the classic answers in the form of links, Izvestia found out. In May, Google actually began the largest search service transformation in recent decades. The corporation is gradually moving away from the usual model in which the user clicked on links to websites and independently studied the sources. Now AI agents and generative responses are coming to the fore, which independently collect, analyze and retell information.
Some of the Google Search and AI Mode updates begin to be implemented as early as May 26 as part of the Core Update. In the summer, the company also plans to launch information agents capable of tracking news, finance, sports, announcements and social media posts around the clock, and then generating ready-made personalized summaries. At the same time, Universal Cart, Gemini Spark, Google Workspace and other Google ecosystem services will receive updates.
The new model can seriously change the economy of the Internet, experts interviewed by the publication believe. If the user receives a ready-made answer inside the search engine, the need to switch to media sites, blogs, or services is gradually decreasing. This can lead to a noticeable drop in traffic for traditional media and author sites. At the same time, Russian companies are not going to copy Google's approach and abandon the classic search results.
Zen believes that AI should remain a tool for editorial offices, not a substitute for journalism. Alexander Tolokolnikov, managing Director of the platform, recalled that in March 2026, the service launched the Glyph AI assistant, trained on more than 200 million news materials. It helps users navigate the agenda and generates personalized digests. At the same time, the platform is developing media tools, including technologies for automatically creating headline options for different audience segments.
In addition, Zen is preparing to launch an AI studio for editorial offices, which will automatically convert one material into several formats — articles, posts and other types of publications. At the same time, the company emphasizes that the basis of any AI systems is still the original content created by journalists and editorial offices, said Alexander Tolokolnikov.
VK relies on personalization and promotion of copyrighted materials, said Denis Shaveynikov, head of the Search department at AI VK. According to him, the company actively uses vector search and multimodal neural networks.
— Algorithms analyze not only text, but also images, video sequences and other elements of publications, which allows you to better understand the meaning of the content and more accurately compare it with queries. At the same time, AI in VK services helps to promote original copyrighted materials, including small authors," he said.
In Rambler & They also talk about the gradual transformation of the search model: it becomes more convenient for users to ask questions directly to AI assistants instead of manually collecting data.
— Services are gradually shifting from classical search to systems that help people get ready-made information and personalized answers faster. However, the final choice of the content consumption model will still be left to users," said Anna Ivanova, Executive Director of the Rambler portal and director of the company's Content and technology department.
According to Mikhail Slivinsky, the ambassador of Internet platforms at Yandex Search, today it is important not just to find a link, but to quickly understand the topic, compare options and make a decision.
— The system analyzes the user's intention and selects the most useful response format — classic output, Alice's quick response, dialog mode or thematic search. At the same time, AI answers are shown only in cases where they really help solve the problem faster, and links to websites continue to be an important part of the ecosystem," he noted.
He also stressed that Alice's quick responses already cover about 36% of requests. Such results are generated based on the highest quality and expert materials on the Internet, and website owners can track the presence of their resources in AI responses through Webmaster tools, the expert added.
Hidden Risks of AI Responses
The spread of AI can seriously change the very principle of information consumption on the Internet, according to IT expert Sergey Pomortsev. According to him, the new search format makes the Network much more convenient: a person no longer needs to independently browse dozens of links, compare sources and analyze large amounts of data — the neural network does this automatically and outputs a ready-made result.
However, this model simultaneously creates new risks: the user does not receive a set of different points of view, but an already formed interpretation of the information, which may depend on the specifics of the algorithms, limitations of training samples, or the editorial policy of the platform. As a result, AI is gradually becoming not just a search tool, but a full-fledged intermediary between humans and information, he noted.
According to him, the work of generative systems is basically impossible without original journalistic content. Neural networks do not create primary content on their own — they recycle materials written by the media, experts, and authors. If the media industry starts to lose its audience and revenue due to falling traffic, this may affect the AI systems themselves, as there will be fewer high-quality sources, Sergey Pomortsev emphasized.
— Today, foreign AI platforms have learned to identify users not only by IP address. Digital fingerprints of devices, login history, payment information and other indirect signs are used. Therefore, circumventing the restrictions is becoming increasingly difficult. If we talk about numbers, the platforms themselves practically do not disclose official statistics," explained Mikhail Shurygin, chairman of the ROCIT Commission on Cloud Technologies and Information Security.
The Internet is gradually ceasing to be a single space. Large foreign platforms are increasingly operating under political, sanctions and regional rules. And the main conclusion here is not even about a specific resource. According to him, it is impossible to build critical business processes on technologies, access to which can be disabled at any time by the owner of the Internet service.
Igor Baranov, an expert in the field of digital technologies, notes that Google is no exception — a similar transition to AI search is gradually taking place in almost all of the world's largest ecosystems.
— Microsoft's Bing system is actively moving along Google's path, which also relies on generative responses and built-in AI assistants. The largest technology platforms are trying to keep users inside their own ecosystem for as long as possible, minimizing the need to switch to external sites," he explained.
Against this background, according to the expert, interest in alternative and private search engines such as DuckDuckGo, which focus on privacy and try to avoid excessive personalization of search results, may grow.
As Igor Baranov noted, the strengthening of AI algorithms inevitably increases the risk of the formation of a so-called filter bubble, when the user sees only the information that the algorithm considers relevant to him. In the future, this may lead to an even greater fragmentation of the information space and a decrease in the diversity of opinions on the Internet.
Переведено сервисом «Яндекс Переводчик»
