AI DRIVEN QUALITY ASSURANCE AUTOMATION IN SOFTWARE TESTING: A FRAMEWORK AND IMPLEMENTATION

Quality assurance (QA) is critical in software development, yet manual QA is time-consuming and error-prone.Recent advances in AI have enabled more intelligent, automated testing, significantly improving defectdetection and efficiency[1][2]. This paper presents an AI-augmented QA framework that integrates machinelearning (ML)–based defect prediction, generative test-case creation, and explainable analysis into the QApipeline. The proposed system uses historical test data to train predictive models that prioritize test scenarios,while generative models create new tests for uncovered code paths. A feedback loop continuously retrains theAI from test outcomes. We implement a prototype in a CI/CD pipeline (using Python, Jenkins, and Selenium) todemonstrate the concept. Experimental observations (inspired by related work) indicate higher bug detectionrates and reduced manual effort. Our results suggest that AI-driven QA can achieve near “expert-level”coverage and speed[3][4]. This implementation-driven approach highlights how AI can automate and optimizeQA processes in practice.