FortiGuard Advanced Bot Protection employs continuous machine learning and behavioral analysis to protect applications from account takeover attempts, data theft, and service disruption. A continuous feedback loop allows the creation and then detection of emerging attack patterns and bot behaviors. The service is able to distinguish between human users, legitimate bots (chatbots), and dangerous bots.
Browser fingerprinting detection
FortiGuard Advanced Bot Protection can block or limit simple bots by IP or signature. To detect more sophisticated bots, the service analyzes crawler-specific attributes and checks for inconsistencies in the browser and operating system. Not only does it follow every unique user agent that transacts with the application, but it also feeds the algorithm with new patterns.
Biometric-based detection
For more sophisticated bot threats, FortiGuard Advanced Bot Protection monitors, collects, and analyzes client events. These include mouse movements and keyboard clicks to cross-validate with identification and intent-based parameters. It also looks for CAPTCHA bypasses and challenges these bots, to eventually provide a risk score for the organization to act upon.
Continuous, adaptive learning
As bots continuously evolve to bypass security mechanisms, new bot behaviors and evasion techniques must be learned and applied to defenses. This solution uses deep learning and data correlation between multiple dimensions, challenging and probing variations over time to provide accurate risk evaluation. It also accurately classifies good bots.
