As is known, credit risk assessment in financial institutions is given important attention. Credit risk assessment is of great importance both at the stage of credit application consideration and in the process of credit monitoring. Traditionally, banks use expert assessments and scoring models based on mathematical statistics to assess credit risk. However, with the development of technology, financial institutions are increasingly turning to AI to optimize business processes. AI has become an integral part of the financial industry, transforming decision-making processes and reducing the risks of mistakes. It is believed that the most common use of machine learning (ML) in the financial sector is scoring. Machine learning in credit scoring has advantages over traditional scoring. ML algorithms can analyze big data and predict the probability of a loan repayment more accurately. AI can process credit applications faster and more efficiently than a credit manager. ML algorithms make decisions based on sound mathematical models and exclude the subjective people factor. In this study, the author considers and compares approaches to credit risk assessment based on the traditional method and the method using AI and ML algorithms. The goal of the study is to determine the accuracy of the forecast of the probability of loan default by borrowers based on mathematical models, using AI and without using AI. The novelty of the study lies in the fact that the author, together with the bank's employees, conducted an experiment to assess the credit risk of borrowers using ML algorithms and identified the advantages of this method.