Gargoyle Guard: Enhancing Cybersecurity with Artificial Intelligence Techniques

Common user authentication methods have limitations that pose significant cybersecurity risks today. Users may inadvertently or intentionally leave their devices and sensitive applications vulnerable to attackers. Although timing-based and dedicated hardware solutions reduce this risk, they do not guarantee foolproof user authentication. To address this security gap, we introduce realtime user activity fingerprinting. This technique continuously verifies users based on their unique behavior patterns. In this paper, we introduce Gargoyle Guard, a smart software solution, and provide insights into our comprehensive software development process, including requirement analysis, software architecture, and design. Gargoyle Guard employs a machine learning model we developed, trained, and tested to dynamically evaluate its authentication system's performance, error rates, and accuracy.