<![CDATA[The growing complexity of risks and crisis situations across various sectors has encouraged the development of Artificial Intelligence-based Decision Support Systems (DSS) to support more accurate, responsive, and data-driven decision-making. These systems are considered valuable because they can help decision-makers analyze uncertainty, recognize potential threats, predict possible impacts, and determine appropriate actions in complex conditions. However, the use of black-box AI models creates serious challenges, particularly regarding transparency, accountability, interpretability, and user trust. Therefore, this study focuses on examining the trend of Explainable Artificial Intelligence (XAI) integration in DSS for risk and crisis management, identifying the XAI methods commonly applied, and revealing research gaps that still need further attention. The study applies a Systematic Literature Review (SLR) method using the PRISMA approach, involving 47 selected articles obtained from Scopus, IEEE Xplore, ScienceDirect, and Google Scholar databases, published between 2020 and 2025. The findings show that XAI plays an important role in improving transparency, interpretability, and trust in AI-based DSS, with SHAP and LIME being the most frequently used methods. Nevertheless, gaps remain, especially limited XAI implementation in real-time crisis scenarios and insufficient human-centered design approaches. This study contributes a conceptual framework integrating DSS, XAI, and risk management.]]>
Copyrights © 2026
