The National Bank of Ukraine is creating a prototype of an AI agent system that replicates the country’s population structure to quickly detect changes in citizens' inflation expectations. This tool allows the regulator to check society’s reaction to economic events even before conducting real sociological surveys.
The National Bank of Ukraine is creating a prototype of an AI agent system that replicates the country’s population structure to quickly detect changes in citizens' inflation expectations. This tool allows the regulator to check society’s reaction to economic events even before conducting real sociological surveys.
This is stated in a study by National Bank specialists published in the "Inflation Report" for October 2026.
The development of AI agents was a response to the high volatility of Ukrainian sentiment during the war. Virtual agents receive up-to-date macroeconomic data and information about the tone of news in the Ukrainian media space. It is noted that in September 2026, synthetic surveys based on artificial intelligence tools helped the NBU identify the risk of worsening inflation expectations in a timely manner, which was later confirmed by real sociological data.
In addition to modeling population response, the NBU is expanding the use of AI and Big Data for other analytical tasks:
In the summer of 2026, after massive Russian strikes on logistics, the AI assistant collected a registry of strikes, classified 335 components of the consumer price index, and built a scenario model in a few days. In addition, the language model processed more than 10,000 records from ACLED’s database of Russian attacks.
The NBU monitors search queries and reactions on social media. For example, the analysis showed that the news about the introduction of a 2,000 UAH banknote did not cause concern among citizens, unlike the negative background surrounding fuel, food, or transportation prices.
The use of alternative data (online prices, satellite images of illumination, vacancies) allows us to estimate inflation two weeks earlier than the SSSU with an average error of only 0.01 pp.
The National Bank emphasizes that despite the introduction of synthetic surveys and automated modeling, the role of a human analyst remains crucial. NBU specialists control algorithms, check data sources, and interpret results to separate real economic signals from information noise.
