Software testing is an important step in software development as it ensures the reliability, functionality, and performance of developed software. However, manual testing approaches are often labor-intensive, time-consuming, and prone to human error. Automating this phase offers significant benefits, making it a key area of interest for researchers. This study aims to explore the adoption of generative AI (GenAI) technologies in automating software testing and provides a thorough review of recent advancements. This research work includes a comparative analysis of existing studies, revealing a predominant focus on Large Language Model (LLM)-driven solutions for generating the test cases, with notable applications in unit testing and integration testing. Evaluation metrics such as test coverage, efficiency, and relevance are frequently used to assess the effectiveness of these approaches. Furthermore, the review identifies critical limitations in current GenAI-based solutions and suggests potential directions for future work. This work offers a timely contribution to the software engineering community, laying the groundwork for more advanced research work in the domain. © 2025 IEEE.

https://www.scopus.com/pages/publications/105017009571?origin=resultslist