Loan scoring and fight against swindlers: AI in Russian banking sector 14.03.2019
Loan scoring and fight against swindlers: AI in Russian banking sector

The integration of artificial intelligence into the banking sector positively affects both financial market players and their customers. AI systems in credit institutions reduce the impact of human factor on decision making, promptly analyze information, and automate communication with users. At the same time, bank clients obtain services of higher quality. The article describes in detail the application of artificial intelligence in the Russian banking system.

Banking scoring

Loan granting banks use scoring technology. It is referred to a system that financial institutions apply to assess a borrowing capacity of people. This program is filled with various information about potential borrowers:

  • age and gender;
  • family status;
  • income level;
  • length of employment;
  • credit history;
  • place of work and education;
  • availabilityofcars,etc.

Each criterion is assessed in a certain amount of scores, which are eventually summarized. Such an inspection allows banks to determine the financial responsibility of people. Afterwards, they decide whether to issue a loan or not.

Today, financial organizations are integrating innovative scoring solutions into their operations. It is referred to AI systems helping to analyze information about bank clients using artificial intelligence and machine learning. They examine data from social media and information analysis bases, analyze the activity of potential borrowers on the Internet, and even take into account their behavior.


Artificial intelligence at Russia-based Tinkoff Bank automatically analyzes customer data and determines an appropriate lending limit. Afterwards, the financial organization keeps examining borrower’s expenditures, habits, and other factors. This information is used to correct the amount of a loan.

LIME has developed an innovative scoring system that analyzes up to 10,000 consumer characteristics in several seconds. The program relies on both person’s credit history and behavior. It learns how individuals fill out the form, how much time they spend on each action, as well as considers other behavioral features. Such an approach makes a loan granting process automated.

In 2016, Mail.Ru Group and Equifax Credit Bureau have designed their own scoring system for banks. Their software analyzes big data. As the result, it defines reliable clients among users who have been refused a loan for various reasons (for instance, because of a poor credit history). This kind of product opens a wider lending market for banks.

Fight against swindlers

Artificial intelligence technologies help financial market players to combat fraudsters who try to obtain credits or conduct money transfers by deceitful means. Banks utilize a biometric analysis to identify customers and recognize speculators. Now, AI systems already know how to determine people by voice, face, and even speed of typing.


Russia applies remote identification technology based on biometric data at Sovcombank, VTB, Pochta Bank, and other lending agencies.

Tinkoff Bank’s computer vision recognizes documents with the accuracy of 90%. Thistechnologyisalreadyusedtoidentifycustomers’ faces. Due to this solution, one has managed to reduce false document frauds sixfold.

A Swedish company, BehavioSec, has developed an AI system able to identify people by speed of typing, keystroke force, and other aspects. Such a biometric analysis allows to recognize a swindler who tries to take on a different persona.

AI Conference:  Loan scoring and fight against swindlers: AI in Russian banking sector 1


According Gartner’s forecasts, the majority of communication processes involving entrepreneurs and their clients will be automated by 2020. Chatbots will help them to implement the concept.

Currently, virtual assistants are increasingly replacing human call center operators in the banking sector. They provide users with information about goods and services, give tips, show an exchange rate, and support in transferring funds from card to card.

Chatbots are already used by the following major companies: Bank of America, Capital One, American Express, DBS, and Master Card. Russia has integrated virtual assistants into Sberbank, VTB, Tinkoff Bank, Alfa-Bank, and other depository institutions.


In 2016, Russia’s Tochka Bank launched a chatbot on Facebook messenger. It enables clients to examine their account information, find ATMs and cash desks, as well as carry out payments.

Tinkoff Bank has several chatbots. For example, the bank launched a virtual assistant called MoneyTalkBot on Telegram messenger in 2015. It is designed for transferring money to other users. In 2018, the financial organization added a bot for businesspeople. It is equipped with educational and entertainment features.

Besides, all Tinkoff Bank requests are processed by special virtual assistants. 20% of all dialogues with customers are automated and conducted without operators.

AI systems for debt collection

Financial market insiders apply artificial intelligence technologies to grapple with debtors. AI systems inform clients of debts, detect users requiring a reminder of payment, and help collection agencies to take an individual approach to every consumer.


In 2018, VTB launched an automatic system for collecting loan debts among bank customers. This software automatically informs users of payment delay, amount of debt and time limits, as well as specifies when it should be paid.

Russia-based Binbank is building self-learning uplift models. They allow to draw up a list of individuals who need advice and compulsory reminding of loan payments. They also define consumers for whom calls and notifications of indebtedness will be inefficient.


Artificial intelligence is one of the most promising technologies in the banking sector. Nowadays, AI systems help financial establishments to have a better idea of consumers’ need, enhance customer experience, and increase revenues. AI platforms allow bank customers to obtain more personalized services and promptly solve relevant issues.

AI-industry news and cases for you