AI-Based Self-Healing Automated Testing Framework for Web Applications

Rapid changes in modern web applications frequently affect automated testing systems because traditional scripts depend on fixed locators and static page structures. Even small interface modifications may cause automated test execution to fail although the actual functionality remains unchanged. This research proposes an AI-driven self-healing automation framework capable of identifying failed elements, analyzing updated DOM structures, and locating substitute elements through intelligent similarity analysis. The framework automatically repairs broken locators during runtime and allows execution to continue without human intervention. The proposed solution improves test stability, lowers maintenance requirements, and supports uninterrupted execution inside Continuous Integration and Continuous Deployment (CI/CD) environments. Experimental observations indicate that the framework significantly decreases false failures and enhances the overall efficiency of automated web testing.