Applications of Artificial Intelligence in Disaster Management: Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Abstract

This study presents an examination of artificial intelligence (AI) in the disaster management (DM) cycle to support human decision-making systems. There is an urgent need for holistic application strategies that support disaster risk reduction given the increasing frequency and severity of disaster events. This stems from the growing adoption of AI technologies, including machine learning (ML), deep learning, and natural language processing in transforming disaster management. This is to enable the development of high-performance tools for assessing disaster impacts and addressing un-met human needs. This study reviewed journals and conference proceedings articles from three different academic databases: Web of Science, PubMed, and IEEE Xplore. Article selection and query refinement were conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to support methodological approach. The review explored scholarly perspectives on AI application in DM phases to support human-centered decision-making processes in reducing disaster impacts on human lives. Findings from the review indicate that most existing studies focus primarily on the AI application in the response phase. Despite its potential, integrating AI into disaster management remains challenging, particularly due to widespread misconceptions about adoption of AI, as well as ongoing ethical, trust, and security concerns. This study emphasizes the comprehensive application of AI to enhance planning, encourage proactive measures, and improve response and recovery efforts during disaster events.