In social networks, AI agents predicted the role of economic entities

Russian-language social networks are actively discussing the prospect of an independent AI agent economy in which digital assistants will become separate entities. This trend was recorded by a study by Brand Analytics, commissioned by VTB as part of the Data Fusion conference. In social networks, users are already raising questions about the risks, monetization models, and regulation of the new system, the researchers note.

The authors of the paper "Perception of AI agents through the eyes of social media users" analyzed the mentions of the term in various word forms in Russian. The most popular topic turned out to be practical application — 44% of the messages. The authors discuss ways to transform AI from a "talking" tool into a "doing" one and create a "personal Jarvis" — a digital assistant acting on behalf of a person.

The study captures a new role for humans: they become the orchestrator of several AI agents. Formulates goals, allocates tasks, and controls processes. However, managing multiple tools in parallel carries risks of cognitive exhaustion.

The role of a person as an "agent manager" is a logical but intermediate stage, the study notes. "Now advanced users are forced to coordinate several AI tools in order to get the maximum effect from using the technology. However, in the future, the architecture of agent systems itself will be simplified.: There will be meta-agents capable of coordinating other agents and assuming a significant part of operational control," predicted Vadim Kulik, Deputy President and Chairman of the VTB Management Board.

According to him, the key barrier to a full—fledged digital assistant is the ability of AI agents to reliably deliver tasks. "What matters to the user is not how smart the assistant is, but whether, for example, they can make a payment, select a financial product, arrange a service, or optimize costs without unnecessary steps," explained the VTB top manager.

The study captures the formation of a hypothesis about a new market where digital agents will become independent economic entities. "As a bank that actively invests in the development of AI agents, we see that the beginnings of the digital agent economy are already emerging. But it will take several more years to achieve a full—fledged system with stable rules and regulations," Kulik said. In the future, AI agents will be able to independently select products, optimize costs and manage financial flows, he added.

35% of discussions are devoted to the perception of AI agents, the study notes. At the same time, user opinions range from complete admiration to deep skepticism. In these opinions, 20% of the discussions are devoted to the fact that AI agents are becoming the norm and an integral part of professional and daily activities. At the same time, some users (11% of discussions) express distrust or disappointment in AI assistants due to high expectations and aggressive marketing. Technological issues — architecture and choice of platforms — occupied 7% of the discussions. Philosophical discussions about AI consciousness and the concept of general artificial intelligence accounted for 3%.

"The polarization of perception of AI agents is an expected stage of maturity of any new technology. Skepticism is also important for the market. He points to the gap between promises and user experience, which sets the vector for product development and more honest communication," Kulik emphasized. He also warned that AI can become a tool for fraudsters, so in critical areas, including finance, it is necessary to use the services of large trusted players.

All important news is on the Izvestia channel in the MAX messenger.