USE OF ARTIFICIAL INTELLIGENCE IN SECURITY SYSTEMS OF E-COMMERCE SERVICES

Abstract

The article considers artificial intelligence based approaches for improving existing security systems ofe-commerce services. Along with the constantly growing number of transactions passing through e-commerce platforms, thefield of opportunities for malicious attacks also expands, as adversaries more and more often use and refine automated systemsthat imitate the behavioral patterns of real users. Traditional security solutions that rely on signatures and static threshold mechanisms adapt poorly to the dynamic change of attackers tactics and therefore require further development. Modernprotection systems that employ artificial intelligence suggest a move from static checks to continuous, real-time analysis ofevent streams. This approach makes it possible to observe user interactions with system in detail and to respond to threats in a timely manner. Industry leaders continuously develop this direction and set tendencies that later become common standards for implementing the security component of e-commerce services. The article also outlines the advantages of behavioralbiometrics, which allows modeling individual user characteristics and building stable profiles for accurately distinguishinglegitimate sessions from unwanted ones. The use of continuous machine-learning and analytical methods increases the speedof anomaly detection within the event flows occurring in the system.Methodology. The article presents an analysis of modern methods of using artificial intelligence systems that can be applied to the development of a multi-layered security system architecture. Scientific novelty. The paper summarizes modern methods of applying artificial intelligence systems to counter currentsecurity challenges faced by the protection systems of e-commerce services.Conclusions. The integration of artificial intelligence technologies with existing security systems of e-commerce services significantly expands their ability to adapt to modern threats. Applying artificial intelligence with the support of continuousdata-analysis models creates a balanced strategy for protecting e-commerce services, improves the accuracy of detectingunwanted activity, reduces financial losses for service providers, and strengthens the trust of end users in e-commerce services.