Transforming Software Testing with AI-driven Test Automation : Exploring Implementation, Challenges, Industry Perceptions, and Future Impact on Software Quality

Sammanfattning

Today's software development faces many challenges, particularly in terms of software quality, which is affected by frequent updates, diverse user environments and complex dependencies. Testing plays a critical role in ensuring quality, but manual testing is often time-consuming and at odds with today's fast-paced development cycles. Test automation addresses some of these issues by increasing efficiency and reducing repetitive work. However, as the number of automated tests increases, so does the maintenance burden. Organisations are exploring the use of artificial intelligence to address these challenges. This paper explores how Artificial Intelligence (AI) is currently being used in test automation, its key advantages and limitations, and the current perceptions of professional developers and testers in the field. Using a qualitative approach, six semi-structured interviews were conducted with participants from areas related to testing and AI, which was used as empirical data. Thematic analysis was used to analyse the data collected from the interviews, and the results showed a variety of application areas for AI in test automation, with benefits of speed, scalability, productivity and creative support, but also highlighted the limited understanding of AI, human trust issues, legal and ethical concerns, and unreliable results. Perspectives on the future use of AI in test automation varied, highlighting the uncertain yet stable aspects of AI’s future development, its role in test automation, human control, responsibility, and impact on the profession of testing.