AI-Based Automated Defect Detection and Test Prioritization in Software Quality Engineering

Abstract

The increasing complexity of software systems has created a need for quality engineering approaches that can identify defects efficiently and allocate testing resources according to risk. AI-based automation offers a mechanism for transforming conventional test execution from a largely rule-driven process into an adaptive decision-making activity. This paper develops a conceptual framework for AI-based automated defect detection and test prioritization by synthesizing principles from the provided literature on graphical authentication, password usability, user behavior, and security-oriented evaluation. Although the cited studies primarily address authentication rather than software testing, they collectively demonstrate the importance of usability, behavioral variability, attack resistance, user choice, and systematic evaluation in automated security-sensitive systems. These principles are mapped to defect detection and test prioritization through a proposed pipeline consisting of test-data acquisition, defect-feature extraction, risk scoring, intelligent classification, prioritization, execution, and feedback-based model refinement. The analysis argues that AI-driven test automation can improve the allocation of testing effort by prioritizing cases associated with higher predicted defect probability or security impact. The framework also incorporates the broader perspective of AI-driven test automation frameworks for modern software quality engineering (Ramamurthy, 2023). The paper concludes that AI-based prioritization should not replace conventional testing controls; rather, it should function as an adaptive decision-support layer whose effectiveness depends on representative data, explainability, continuous validation, and careful management of false-positive and false-negative outcomes.