Software Testing with Large Language Models: An Interview Study with Practitioners

TL;DR

This study explores how software testing professionals use large language models in practice, revealing iterative workflows, the importance of human oversight, and the need for structured guidelines to support their integration into testing processes.

Contribution

It provides the first practitioner-informed qualitative insights and preliminary guidelines for integrating large language models into software testing workflows.

Findings

Testers use iterative prompt engineering and evaluation.

Human oversight is crucial due to LLM limitations.

Practitioners emphasize cautious, reflective use of LLMs.

Abstract

\textit{Background:} The use of large language models in software testing is growing fast as they support numerous tasks, from test case generation to automation, and documentation. However, their adoption often relies on informal experimentation rather than structured guidance. \textit{Aims:} This study investigates how software testing professionals use LLMs in practice to propose a preliminary, practitioner-informed guideline to support their integration into testing workflows. \textit{Method:} We conducted a qualitative study with 15 software testers from diverse roles and domains. Data were collected through semi-structured interviews and analyzed using grounded theory-based processes focused on thematic analysis. \textit{Results:} Testers described an iterative and reflective process that included defining testing objectives, applying prompt engineering strategies, refining prompts, evaluating outputs, and learning over time. They emphasized the need for human oversight and careful validation, especially due to known limitations of LLMs such as hallucinations and inconsistent reasoning. \textit{Conclusions:} LLM adoption in software testing is growing, but remains shaped by evolving practices and caution around risks. This study offers a starting point for structuring LLM use in testing contexts and invites future research to refine these practices across teams, tools, and tasks.