Digital Employees in Banking: AI Assistants Take on More and More Functions

From Chatbots to AI Agents

Under the umbrella of the Association FinTech, pilot projects are underway to deploy AI agents built on domestic models. Where artificial intelligence was once limited to simple chatbots answering basic customer questions, banks are now working with full-scale AI agents. These systems can analyze customer behavior in mobile banking apps, assist staff with document verification and compliance, generate hypotheses for new products, and forecast demand. One participant in the Association FinTech initiative, OTP Bank, is already using such agents to optimize digital services across key business areas.

The project, which involves 15 Russian banks, aims to develop a shared methodology and baseline architecture for AI solutions applicable across the entire banking sector. For banks, this promises efficiency gains; for domestic IT companies, sustained demand for locally developed GPT-style models; and for the broader economy, faster digital transformation combined with lower operating costs.

A New Level of Customer Experience

The potential impact of AI agents is significant. They can fundamentally reshape customer service, from delivering real-time personalized recommendations to instantly processing requests without human involvement. For banks, this means not only easing the workload on staff but also accelerating the rollout of new digital products. Market analysis, evaluation, and hypothesis generation that once took weeks can now be completed in a fraction of the time.

Together with our partners, we successfully confirmed the hypothesis that Russian large language models and tools from domestic vendors are suitable for building AI agents and can be effectively applied by financial organizations across different processes. Over the course of the project, a number of use cases were implemented that organizations can further develop. The most successful results were uploaded to the AI Laboratory and will be used by other members of the Association

Alexey Sidoryuk

AI Advisor to the CEO of the Association FinTech

For now, deployment is focused on the domestic market. However, successful implementations could eventually serve as a foundation for exporting such solutions to friendly countries with similar regulatory environments.

From Early Experiments to Systemic Solutions

Russia’s financial sector has been at the forefront of digitalization. Banks were among the first to apply AI to routine tasks such as document processing, first-line customer support, and transaction analysis. Over time, these capabilities expanded.

A key inflection point came when Sberbank launched its own large language model, GigaChat, as part of a broader AI ecosystem, setting a benchmark for the industry. In parallel, the Central Bank of Russia began working on measures to expand access to data for model training.

Today’s AI agent pilots mark a shift from fragmented solutions to a systemic architecture in which AI becomes an integral part of a bank’s digital environment. Solutions that pass selection will be added to the AI Laboratory knowledge base and made available to all Association FinTech members.

Forecasts and Strategic Implications

AI agents are emerging as a tool for boosting business efficiency with measurable economic impact, not only in banking. Insurance companies are already using AI for risk assessment, while investment platforms rely on it for personalized recommendations. By 2027–2028, digital employees are expected to be active both in customer-facing services and in internal operations.

Russia’s digital economy is increasingly built on intelligent algorithms developed by domestic companies, with the banking sector setting the pace.