PrivatBank uses artificial intelligence to generate about 80% of personalized offers for customers. The algorithms help the bank determine which discounts, cashback, promo codes, and partner promotions may interest a specific user, replacing mass offers with individual recommendations.
This was announced on November 18 by Serhii Kyriukhin, head of the unsecured lending department at PrivatBank, during the "Consumer lending: how banks compete for customers on Black Friday" roundtable, organized by "Financial Club."
According to the bank representative, large datasets are used to generate recommendations. Artificial intelligence models are trained and adjusted daily to identify the audience for whom a particular offer will be relevant now or in the near future.
One of the main applications of such technologies has become the "Pryvit" loyalty program in the Privat24 app. Through it, the bank and its commercial partners offer customers cashback, discounts, and special purchasing conditions. Artificial intelligence helps tailor these offers based on the anticipated needs of users.
Instead of the same set of promotions for all customers, the system generates individual recommendations. This allows the bank to reduce the number of irrelevant offers, and enables retail companies to reach an audience potentially interested in their goods and services.
PrivatBank has previously disclosed the scale of its technology use. In December 2025, Inna Turliun, head of the "Pryvit" loyalty program, reported that the platform unites more than 15 million active users, and its artificial intelligence system generates about 735 million personalized recommendations annually.
According to her, the algorithms allow analyzing long-term patterns of purchasing behavior and forecasting the customer's interest in certain goods even before direct demand arises.
In one example provided by the bank, AI-based recommendations attracted over a million customers to partner promotions. More than half of them made purchases, which, according to PrivatBank's assessment, also contributed to an increase in the average check.
At that time, more than 15,000 commercial partners were connected to the program, including major retail chains "Foxtrot," "Aurora," and WOG. For businesses, the bank has also developed a marketing dashboard in Privat24, allowing the launch of promotions and targeting specific customer categories independently.
In April 2026, PrivatBank board member Dmytro Musiienko stated that 87% of "Pryvit" program users receive unique sets of offers generated based on data analysis.
This figure differs from the now-mentioned 80%: the first characterizes the share of users receiving individual sets, while the second is the share of personalized offers generated using artificial intelligence.
PrivatBank applies machine learning not only in marketing but also in credit risk management. According to Musiienko, technologies are used in risk assessment and issuing unsecured loans. Among other development areas, the bank has mentioned creating digital financial assistants, automating internal processes, and detecting potential fraud.
At the same time, the statement about forming 80% of offers with AI does not mean that artificial intelligence independently makes the same share of credit decisions. This primarily concerns personalizing offers to customers.
During the roundtable, Serhii Kyriukhin also shared expectations regarding the consumer lending market in 2027. According to his forecast, there will be no significant downturn, but the growth rates of loan portfolios are unlikely to exceed 2026 figures.
The PrivatBank representative noted that further lending development will depend on the security and energy supply situation, employment and income levels of the population, and the quality of loan portfolios. The latter factor affects banks' willingness to extend new loans and the terms offered to borrowers.