AI-Powered Software Testing Tools: Full Autonomy Remains a Distant Goal

AI-powered software testing tools have taken test automation to a new level, promising improved efficiency, reduced 
maintenance effort of test-automation code, and possibly even enhanced defect-detection. In this paper, we systematically 
review and classify the set of 56 AI-assisted software testing tools available on the market (as of 2024). 
Our review shows that AI-powered software testing tools can provide benefits to software test engineers, by potentially 
increasing effectiveness and efficiency of their software testing tasks, from test planning, to test-case design, to test 
execution and test-code maintenance. However, there are limitations and possibly even issues in these tools—such as false 
positives, and lack of domain (contextual) knowledge. A human test engineer should still peer review and validate any testing 
artifacts, e.g., test cases and test code, generated by the AI tools. For years to come, we predict that humans and machines 
(software testers and AI-powered testing tools) have to still work together. The road for full autonomy of these tools is still 
long.