Advancing Continuous Authentication Using Smart Real-Time User Activity Fingerprinting

Continuous authentication enables a system to continuously verify a user's identity throughout an active session, eliminating the need for a one-time login. One effective approach to continuous authentication is real-time user activity fingerprinting. It analyzes a user's behavior, including keystroke patterns, mouse movements, and browsing habits, to create a unique user profile or “fingerprint.” Real-time user activity fingerprinting offers multiple advantages over traditional authentication methods like passwords or tokens. It enhances security by making it difficult for attackers to imitate a user's behavior, while also providing the convenience of eliminating the need to remember passwords or carry physical tokens. In this paper, we present Gargoyle Guard, an intelligent software solution that we design, develop, and integrate. We also provide a comprehensive software development methodology encompassing requirement analysis, software architecture, and design. Gargoyle Guard utilizes a machine learning (ML) model that we develop, train, and verify. This ML model dynamically assesses the performance of the smart continuous authentication system, determining error rates and accuracy factors.