NEURAL NETWORK TECHNOLOGIES AS A TOOL FOR DATA PROTECTION IN THE BANKING SECTOR

Abstract

The article analyzes how neural network technologies and artificial intelligence systems are used to protect data in the Russian banking sector between 2023 and 2025. It is evident that major financial institutions are moving from purchasing ready‑made vendor solutions to creating their own AI platforms and building an autonomous “digital immunity”. Based on official reports from six large commercial banks (Sber, VTB, Alfa‑Bank, T‑Bank, Rosselkhozbank, Gazprombank), a comparison was made of various strategic models – from total automation to in‑depth security testing of IT infrastructure. It is evident that the implementation of multi‑agent systems, a large number of language models (LLM) and machine learning algorithms makes it possible to achieve a detection accuracy of cyber threats of up to 93–99 %, significantly reduce the response time to a problem, and minimize financial losses from fraud and social engineering. The paper demonstrates the effectiveness of a proactive approach to information security for protecting confidential data from threats.