Explainable Artificial Intelligence (XAI) has recently gained attention for making artificial intelligence systems more transparent and trustworthy. However, its adoption within Software Engineering (SE) is still not well understood. This study presents a systematic literature review (SLR) that explores how XAI is being used in SE, along with insights drawn from comparisons with more established domains such as healthcare and finance. In total, 29 primary studies published between March 2020 and April 2024 were analyzed. The findings show that most SE research is based on local explanation techniques, with 86% of the studies using instance-level methods such as LIME. In contrast, global explanation approaches, which provide a broader understanding of model behavior, are used less frequently. Looking at application areas, XAI research in SE is primarily focused on software defect prediction, while other important areas, such as software testing and effort estimation, remain relatively underexplored. In terms of tools, scikit-learn is the most commonly used machine learning library, whereas TensorFlow, PyTorch, and R-based frameworks appear less often. Compared to earlier reviews, this study takes a broader perspective by addressing multiple research questions, examining different types of explanations, and highlighting the use of ML tools. The findings suggest that future work should explore a wider range of XAI methods, make greater use of global explanations, and expand into less-studied SE tasks. Clear reporting of tools and methods is also important to improve reproducibility. Overall, while XAI in SE is steadily growing, it is still at an early stage. Insights from more mature domains may help guide future research and improve the reliability and trustworthiness of SE systems.
