Abstract
Disasters are increasing in frequency and intensity due to climate change, urbanisation, and environmental degradation, challenging traditional emergency management systems. Crowdsourcing — the practice of generating actionable intelligence by mobilising distributed human contributions via digital platforms — has emerged as a transformative paradigm. However, the rapid integration of artificial intelligence has outpaced systematic evidence synthesis. Following PRISMA 2020, we conducted a systematic review of 92 peer-reviewed studies on crowdsourcing in disaster management (2015–2025). Data were extracted using a three-dimensional analytical framework — descriptive, relational, and configurational — and synthesised through frequency analysis, cross-tabulation, and fuzzy-set Qualitative Comparative Analysis (fsQCA). The literature concentrates on floods (27.2%), earthquakes (19.6%), and multi-hazard studies (23.9%), while preparedness (8.7%), mitigation (5.4%), and recovery (9.8%) remain under-studied. Research is geographically concentrated in Global North contexts (35.9% global studies), with Africa, South America, and Central Asia critically under-represented. Automated validation has more than doubled (22.2% to 43.2%), reflecting rapid AI integration. High success is reported in 52.1% of studies, with fsQCA revealing three configurational pathways characterised by multi-layered verification, phased authority integration, and adaptive governance. However, only 15% of studies address digital divide implications, and outcome attribution linking crowdsourcing to measurable disaster risk reduction remains weak. Critically, this review demonstrates that crowdsourcing limitations arise predominantly from institutional resistance, equity blind spots, and coordination failures—not from technological immaturity. Automation has advanced rapidly, yet integration remains predominantly partial, confirming that technology is not the limiting factor. We propose a ten-priority research agenda addressing methodological development, equity and inclusion, ethical AI frameworks, and translational research.