AI applications in disaster governance with health approach: A scoping review

Abstract

Introduction

The increasing frequency and severity of disasters worldwide underscore the urgent need for robust systems that facilitate effective information sharing and decision-making. This study explores the current and potential applications of artificial intelligence (AI) in disaster governance, with a particular focus on health. By examining the transformative capabilities of AI, the study aims to provide practical insights to inform both national and international disaster management policies.

Methods

A scoping review methodology was adopted to investigate the role of AI in disaster management. Systematic searches were conducted in PubMed, Scopus, and Web of Science databases, covering the period from 2000 to 2024. The search strategy employed keywords related to artificial intelligence, disaster management, governance, and health.

Findings

The review identified three core themes where AI enhances disaster governance: governance functions, by improving policy mechanisms, legitimacy, and health system resilience; information-based strategies, through real-time data, predictive analytics, and modeling; and operational processes, by strengthening logistics, communication, and social media management. Together, these applications improve preparedness and response capacity.

Conclusions

This study provides a structured framework for integrating artificial intelligence into disaster governance with a health-oriented approach. By synthesizing evidence across three thematic domains—governance functions, information-based strategies, and operational processes—it highlights how AI can enhance decision-making, strengthen system resilience, and enable more coordinated and equitable disaster responses. These findings offer practical guidance for policymakers and health professionals to develop adaptive, data-driven strategies in the face of increasing global disaster risks.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13690-025-01712-2.

Text box 1. Contributions to the literature
• While numerous studies have explored AI-driven disaster governance with a health-centered approach, no prior research has systematically synthesized these findings to establish an integrated framework. This review consolidates and critically evaluates existing studies, providing a structured analysis of how AI contributes to health-focused disaster governance, with particular emphasis on resilience and policy adaptability.
• Previous research has largely examined AI applications in discrete disaster phases—such as early warning systems, response coordination, and post-crisis recovery—without addressing their cohesive implementation across the entire disaster cycle. This study advances the field by synthesizing diverse perspectives into a unified framework that integrates predictive analytics, algorithmic decision-making, and operational coordination within disaster governance.
• The findings generate novel insights for policymakers and healthcare professionals, illustrating how AI can enhance governance legitimacy, enable anticipatory disaster response strategies, promote equitable interventions, and ensure long-term resilience in public health systems.

Introduction

The rising frequency and severity of disasters—from natural hazards to global pandemics—pose escalating challenges to traditional governance systems. These mechanisms often fall short in managing the complex, rapidly evolving nature of modern crises, particularly due to limitations in processing multi-source, real-time data, which leads to critical gaps in preparedness, response, and recovery [1, 2]. Health outcomes are a vital component of disaster governance, not only because of the immediate threat to lives and infrastructure, but also due to their long-term effects on public health, mental well-being, education, social cohesion, and economic stability [3–6]. Addressing such multifaceted challenges requires the adoption of intelligent, technology-driven strategies. AI-based tools—including crowd simulation systems—are increasingly used to optimize evacuation routes and reduce casualties in high-risk environments [7]. In the field of Environmental Health Risk Management (EHRM), technologies such as Big Data Analytics (BDA) have improved operational efficiency and enabled multi-sector collaboration during emergencies [8, 9]. For instance, Social Big Data (SBD) platforms provide real-time, high-resolution information that enhances situational awareness and supports timely, evidence-based decision-making [10].

Artificial Intelligence (AI) has emerged as a transformative force in disaster governance by automating complex cognitive processes such as forecasting, triage, and strategic planning. Techniques like machine learning (ML) and generative AI—including systems like ChatGPT—have been applied to a wide range of functions, from early warning systems to post-crisis recovery modeling [11, 12]. These technologies can, for example, analyze data from wearable sensors or electronic health records to detect medical emergencies in real time, a capability particularly valuable in resource-limited, disaster-affected regions [13]. Similarly, predictive AI models can anticipate hazards such as floods, wildfires, and landslides by analyzing meteorological and geospatial data, enabling proactive risk mitigation [7, 14]. AI also contributes to situational awareness by analyzing drone footage, satellite imagery, and social media streams, allowingemergency services to respond more rapidly and effectively [13, 15]. Recent advancements underscore the increasing prioritization of digital information governance during public health emergencies. Machine learning techniques—particularly graph-based analytic frameworks—have demonstrated substantial utility in identifying and controlling the dissemination of misinformation, thereby enhancing situational clarity and communication efficacy under volatile conditions [16]. In parallel, a growing body of scholarship has drawn attention to the inherent risks associated with AI integration in healthcare, including algorithmic bias, system unpredictability, and the lack of coherent regulatory oversight. These concerns highlight the urgent need for resilient and ethically grounded governance mechanisms capable of safeguarding equitable health outcomes in disaster contexts [6].

Despite the growing body of research on the application of artificial intelligence (AI) in crisis management, most previous reviews have focused on isolated stages of the disaster cycle or on narrowly defined technical functions. A subset of studies has examined language models in early warning systems [17], while others have analyzed the use of intelligent agents in emergency evacuation simulations [18], and some have assessed damage prediction using multimodal data sources [19]. In the health domain, systematic reviews have shown that AI-based triage—particularly using models such as XGBoost and deep neural networks—offers higher accuracy than traditional methods [20, 21]. Additionally, several investigations have explored reinforcement learning in designing resilient cities [22] and AI-driven frameworks for infrastructure safety and emergency response [23, 24].

Nevertheless, the integration of AI into crisis governance with a health-centered approach remains limited. Many existing strategies suffer from inefficiencies in resource allocation, delayed policy responses, and inequitable access to health services for vulnerable populations [6, 25]. Moreover, AI applications are often fragmented according to hazard type, data source, or infrastructural layer, and few studies have clarified how such technologies can be comprehensively embedded across all phases of prevention, preparedness, response, and recovery. While several integrated frameworks have been proposed [26], most lack explicit alignment between AI-driven technical advancements and the goals of health governance, policy mechanisms, and equity indicators.

This study conducts a systematic analysis of empirical articles across three domains—governance mechanisms, data-driven strategies, and operational systems—in order to propose a conceptual framework that demonstrates how AI can enhance inclusive decision-making, strengthen health system resilience, and reinforce institutional coherence in crisis governance.

Method

Search strategy

This scoping review was conducted following the six-stage framework proposed by Arksey and O’Malley [27], which includes: (1) identifying the research question, (2) identifying relevant studies, (3) study selection, (4) charting the data, (5) collating, summarizing, and reporting the results, and (6) expert consultation. This methodological approach was selected to comprehensively explore the broad and interdisciplinary applications of artificial intelligence (AI) in disaster governance, with a particular emphasis on health. In contrast to systematic reviews or meta-analyses, which typically address narrowly defined research questions or quantitative data synthesis, the scoping review methodology enables mapping a wide range of concepts and identifying knowledge gaps in emerging fields such as AI in disaster management.

1. Research Question: The primary research question guiding this review was: “What are the applications of AI in disaster governance with a health-oriented approach?” This question served as the foundation for search strategy design and inclusion criteria, ensuring the review remained focused on the intersection of AI, health, and disaster governance.

2. Search Strategy: A systematic literature search was performed in three major electronic databases: PubMed, Scopus, and Web of Science. The search strategy employed a combination of keywords. Boolean operators (AND, OR) were applied to effectively combine terms. The complete list of search queries for each database is provided in Table 1.

3. Inclusion and Exclusion Criteria: The following criteria were used to determine study eligibility:

Table 1.

Search queries used for target databases

Databases	Query	Initial results
PubMed	((((emergency [Title]) OR (emergencies [Title])) OR (disaster [Title])) OR (disasters [Title])) AND ((“artificial intelligence“[Title] OR “machine learning“[Title] OR “big data“[Title] OR “internet of things“[Title] OR “expert system“[Title] OR “intelligent system“[Title] OR “Knowledge system“[Title] OR “data mining“[Title] OR “deep learning“[Title]))	558
Web of sciences	((((((((AB=(“Artificial intelligence”)) OR AB=(“machine learning”)) OR AB=(“big data”)) OR AB=(“internet of things”)) OR AB=(“expert system” )) OR AB=(“intelligent system”)) OR AB=(“Knowledge system”)) OR AB=(“deep learning”)) OR AB=(“data mining”) emergency (Title) or emergencies (Title) or disaster (Title) or disasters (Title) AB=(health)	171
Scopus	(ABS ( health ) ) AND ( ( TITLE ( "Artificial intelligence" ) OR TITLE ( "machine learning" ) OR TITLE ( "big data" ) OR TITLE ( "internet of things" ) OR TITLE ( "expert system" ) OR TITLE ( "intelligent system" ) OR TITLE ( "Knowledge system" ) OR TITLE ( "data mining" ) OR TITLE ( "deep learning" ) ) ) AND ( ( TITLE ( emergency ) OR TITLE ( emergencies ) OR TITLE ( disaster ) OR TITLE ( disasters ) )	245

Inclusion criteria: Peer-reviewed articles published in English between 2000 and 2024 that addressed the application of AI in disaster governance, with a particular focus on health-related outcomes, and were available in full text.

Exclusion criteria: Studies were excluded if they did not clearly link AI to both disaster governance and health, or if full-text versions were not accessible.

• 4.Screening and Quality Assessment: Two reviewers (MR and AB) independently screened titles and abstracts against the eligibility criteria. Full texts of potentially relevant articles were subsequently reviewed. Discrepancies during the screening process were resolved through discussion, and a third reviewer was consulted when consensus could not be reached, ensuring methodological rigor and consistency.
• 5.Data Extraction and Charting: A structured data extraction form was used to capture key variables. The extracted data were then coded and grouped into thematic domains to facilitate synthesis. These categories informed the development of a conceptual framework mapping AI applications within the broader context of disaster governance and health systems.
• 6.Expert Consultation: To enhance the validity and relevance of the findings, results were refined through internal discussions within the research team and consultations with a limited number of external experts in disaster management, public health, and artificial intelligence. This iterative process ensured that the synthesis accurately reflected current practices and emerging opportunities in the field.

Results

This section presents the findings of the scoping review in three main stages: (1) search and selection process, (2) development of a conceptual framework, and (3) thematic synthesis of key AI applications in disaster governance. Each part provides a structured understanding of how artificial intelligence is being used—and can further be utilized—in managing disasters with a focus on health and governance.

Study selection process

Fig. 1 illustrates the systematic approach employed for identifying and selecting relevant studies in this scoping review. It outlines the sequential phases, including the initial database search, duplicate removal, title and abstract screening, full-text evaluation, and final selection based on predefined inclusion criteria.

Fig. 1.

Flow chart of the search strategy

The initial database search yielded 974 records. After removing 223 duplicates entries, a total of 751 unique articles remained for title and abstract screening. Based on the predefined inclusion and exclusion criteria, 591 articles were excluded at this stage due to irrelevance or failure to meet the criteria.

Subsequently, 160 articles were selected for full-text review. Of these, 71 studies were excluded after detailed assessment due to either insufficient methodological quality or lack of clear relevance to AI applications in disaster governance within a health context. Ultimately, 89 studies were included in the final synthesis and analysis. (Fig. 1 and Appendix 1 for details of included studies).

The application of AI in disaster governance

To address the complexities of disaster governance, this study proposes a comprehensive framework that categorizes the critical dimensions of AI-driven solutions. As shown in Fig. 2, the framework serves as a foundational model for understanding how AI technologies, such as predictive analytics, Geographic Information Systems (GIS), machine learning models, and the Internet of Things (IoT), contribute to disaster preparedness, response, and recovery efforts.

Fig. 2.

Conceptual framework: existing knowledge and future directions

The framework itself highlights several key dimensions, including goals, health-related disasters, governance areas, disaster management tools, and infrastructures. Following this conceptual framework, the findings are explored in three primary thematic domains: (1) governance functions and elements, (2) information-based approaches, and (3) process and operational methods. These domains encapsulate the diverse applications of AI in enhancing disaster governance with a health-oriented focus.

Integrated framework for AI applications in disaster governance

The framework developed in this study offers a structured and multidimensional approach to understanding the roles and potential of AI in disaster governance, particularly from a health-focused perspective. It is organized into five interconnected dimensions: strategic goals, health-related disaster scenarios, governance areas, disaster management tools and analysis, and infrastructures. Each dimension reflects how AI technologies contribute to strengthening the effectiveness, equity, and resilience of disaster response systems.

Taken together, these five dimensions offer a comprehensive view of how AI can be systematically leveraged to enhance disaster governance across all phases of crisis management. By synthesizing insights from diverse domains, the framework not only fills existing conceptual gaps in the literature but also provides practical guidance for policymakers, health professionals, and technology developers engaged in disaster risk reduction and emergency preparedness.

Strategic goals of AI in disaster governance

At the core of the framework lies a set of overarching goals that define the intended outcomes of AI integration. These include enhancing preparedness, ensuring timely and effective response, supporting sustainable recovery, and improving long-term resilience. AI-driven approaches can support sustainability by aligning with global agendas such as the Sustainable Development Goals (notably Goal 12), while also facilitating the development of smart cities and securing livelihoods in vulnerable settings. Additionally, the framework highlights AI’s potential to strengthen health and social welfare systems by improving access to care, enabling faster crisis response, and maintaining societal stability during and after disasters. Enhancing the performance of governance—particularly through better communication, broader coverage, and more adaptive decision-making—is also emphasized as a key objective.

AI in health-related disaster scenarios

A central focus of the framework is the application of AI in health-related disasters. AI technologies can support various aspects of health crisis management, including epidemic detection, trauma response, and health system optimization. These capabilities are particularly relevant in the context of a wide range of disasters such as earthquakes, floods, wildfires, pandemics, and psychological crises. AI tools can analyze diverse datasets—from environmental sensors to healthcare records—to detect hazards early, allocate resources more efficiently, and manage field operations more effectively. Furthermore, AI can help mitigate the broader social impacts of disasters, including emotional distress, erosion of public trust, and the spread of misinformation, thereby reinforcing public health and safety.

Governance areas and policy functions

The framework categorizes AI applications across key governance domains, underscoring their role in strategic decision-making and equitable service delivery. In the domain of policymaking, AI facilitates the development of evidence-based strategies and real-time decision support systems. It also plays a critical role in addressing social equity by identifying vulnerable populations, monitoring disparities, and ensuring fair distribution of aid and resources. Financial governance benefits from AI tools that support risk modeling, economic impact analysis, and resource prioritization. Additionally, the integration of AI into information governance enables timely risk identification, alert systems, and continuous monitoring, all of which are essential for proactive disaster management. Social governance functions, such as media regulation, public sentiment analysis, and community engagement, are also enhanced through AI-enabled tools.

Disaster management tools and analytical capabilities

One of the most tangible contributions of AI lies in its ability to enhance disaster management operations through advanced tools and analytics. Predictive models, Geographic Information Systems (GIS), the Internet of Things (IoT), and real-time data analytics significantly improve situational awareness and operational decision-making. These technologies allow for the monitoring of spatial and temporal trends, identification of emerging risks, and optimization of humanitarian logistics. The framework emphasizes the role of AI in strengthening institutional capacities for recovery, reconstruction, and environmental protection. Additionally, AI supports the analysis of needs in affected regions, facilitates damage assessment, and contributes to the development of adaptive protocols for crisis management.

Infrastructure and technological ecosystems

Finally, the effective implementation of AI in disaster governance depends on the availability and integration of supportive infrastructure. This includes physical devices such as sensors, surveillance systems, and data-collection tools, as well as digital platforms for big data processing and cloud computing. The framework underscores the importance of robust information and communication technologies (ICT), which enable rapid data sharing and decision support across agencies and jurisdictions. It also addresses the legal and ethical aspects of AI deployment—particularly in terms of data governance, privacy, and accountability—which must be considered to ensure responsible use of emerging technologies.

Thematic analysis of AI applications

The effective application of Artificial Intelligence (AI) in disaster governance requires a multidimensional understanding of how intelligent technologies interact with institutional structures, data ecosystems, and operational practices. This thematic framework synthesizes the core domains through which AI enhances resilience, coordination, and responsiveness in crisis settings. The first domain examines governance foundations—highlighting system architecture, institutional legitimacy, policy frameworks, and healthcare integration—demonstrating how AI supports transparency, trust-building, and adaptive capacity within formal disaster management systems. The second domain focuses on information-based approaches, emphasizing the role of AI-powered analytical tools, digital platforms, economic modeling, and data governance in transforming reactive systems into predictive, real-time decision environments. Finally, the third domain addresses operational and process-oriented strategies, encompassing social dynamics, logistical infrastructures, collaborative coordination, and digital public engagement. Together, these thematic strands offer a holistic perspective on how AI technologies not only automate and optimize disaster response mechanisms but also reshape the structural, informational, and human components of governance in the face of increasingly complex and uncertain emergencies.

Governance functions and elements

Governance functions and institutional mechanisms are foundational to disaster management systems, serving to organize, coordinate, and sustain responses across local, national, and transnational levels. In an era of digital transformation, the integration of Artificial Intelligence (AI), big data, and Internet of Things (IoT) technologies into governance structures redefines how states and institutions prepare for, respond to, and recover from disasters. This category explores the core pillars of disaster governance, including system architecture, legitimacy and authority, public policy management, and healthcare systems. These pillars support adaptive, transparent, and data-driven decision-making processes, ensuring that government actors can enhance service delivery, maintain public trust, and promote institutional resilience—particularly in times of public health emergencies and complex crises. Each subcomponent emphasizes how AI-enabled technologies interact with institutional capacity to reinforce coordination, communication, risk regulation, and equitable access to services.

System architecture

The design and implementation of system architectures play a critical role in enhancing the efficiency, scalability, and resilience of disaster management frameworks. In economically disadvantaged regions, vulnerabilities are often magnified due to the use of substandard building materials and limited financial capacity for infrastructure reinforcement [28]. Systemic resilience in these areas requires integrated architectural approaches that combine traditional knowledge systems with scientific advancements to support sustainable disaster risk reduction [28, 29].

A foundational shift in disaster management architecture has emerged through the concept of AI-infused risk thinking—a systemic framework that embeds artificial intelligence within conventional risk management structures. This approach strengthens overall system safety, facilitates behavioral regulation, and addresses ethical and legal considerations in emergency planning [25]. Additionally, incorporating indigenous expertise within institutional governance structures helps legitimize integrated risk mitigation models and supports sustainable development goals, thereby reinforcing both social and governmental capacities for disaster response [28, 29].

Modern disaster management increasingly relies on layered technological architectures that enable real-time information exchange, advanced analytics, and predictive modeling. Innovations in big data, cloud computing, and the Internet of Things (IoT) have led to the development of dynamic crisis information management systems, which allow decision-makers to draw on historical disaster patterns and localized datasets to forecast health-related impacts and orchestrate timely interventions [30]. Environmental sensors, once used solely for early warning systems, now also contribute to post-disaster spatial and physiological data collection, further enhancing situational awareness and response accuracy [2].

Mobile communication technologies form a core component of resilient disaster communication architecture, ensuring information continuity even when physical infrastructure is damaged [2]. AI-driven decision support systems, integrated into these platforms, promote objectivity in crisis assessment and minimize the risk of biased decisions that may disproportionately affect marginalized communities [31]. The adaptability of such systems hinges on continuous optimization through real-world testing and empathetic, human-centered design [25, 32].

The COVID-19 pandemic highlighted the critical role of legal and institutional structures in shaping the digital architecture of disaster response. Regulatory frameworks, such as those implemented in Hainan Province, enabled secure big data operations, clarified AI accountability, and enforced data protection—essential elements of a robust system architecture [33]. In clinical emergencies, the application of AI has enhanced the operational efficiency of acute care systems, although ethical and adoption-related challenges remain [12].

Integrated architectures also extend to healthcare disaster response. The convergence of big data, cloud platforms, and IoT improves triage systems, resource distribution, and inter-agency coordination. However, the expansion of these capabilities necessitates robust data privacy protections and comprehensive oversight models [30]. Scholars emphasize the need for long-term AI education strategies and the institutionalization of professional training programs to ensure future experts can effectively design, manage, and operate these complex systems [34].

As governments expand digital infrastructure and enhance data-driven governance, researchers and technologists are increasingly empowered to build AI-based applications that reinforce systemic risk mitigation, improve disaster preparedness, and elevate the structural integrity of national and regional disaster management architectures [35].

Enhancing legitimacy and authority

The integration of AI technologies into disaster management has not only improved operational efficiency but has also reshaped the foundations of institutional legitimacy and governmental authority during crises [36]. While these advancements bolster rapid response capabilities, they also demand robust governance structures to uphold public trust, ethical accountability, and transparent decision-making.

To reinforce institutional credibility, governments must prioritize strategic oversight, inclusive policymaking, and dynamic system adaptation [32]. AI-driven decision-support systems contribute to the enforcement of policies and enable timely emergency interventions by accelerating resource deployment and reducing the impact of disasters [10]. The use of interactive data collection tools by local authorities enhances the speed and accuracy of needs assessments and relief distribution, reinforcing the perceived competence of public institutions [37]. Moreover, AI-based fraud detection mechanisms help maintain data integrity, ensuring that governance processes remain reliable and free from manipulation [35, 38].

During public health emergencies, such as pandemics, AI enables healthcare professionals to rapidly identify risk profiles, validate treatments, and streamline clinical trials, contributing to a perception of scientific and institutional efficacy [39, 40]. In parallel, big data analytics and IoT-enabled platforms improve emergency planning and strengthen the legitimacy of governmental responses, particularly in resource-constrained environments [41]. Advanced systems integrating UAV networks further enhance situational awareness, resource coordination, and infrastructure protection—core responsibilities that validate governmental authority during crises [42, 43].

Beyond operational efficiency, disaster governance increasingly requires the cultivation of social legitimacy. Community well-being, equitable access to resources, and transparent communication form the foundation of public trust. AI-enhanced social media analytics allow governments to monitor societal sentiment, adjust policies responsively, and address misinformation in real time [5, 44, 45]. By using AI to counter disinformation and promote factual public messaging, authorities reinforce their role as reliable sources of truth during emergencies [8, 45]. Digital volunteer networks and gamified public engagement strategies further legitimize government efforts by fostering participatory governance and social cohesion [46].

Interdepartmental AI-enabled communication models facilitate coordinated, data-driven responses and transparent cross-sectoral collaboration [1, 4]. Addressing emotional volatility and information asymmetry during crises requires empathetic leadership supported by advanced technologies. Such approaches enhance institutional legitimacy and enable governments to act with authority while maintaining the trust and cooperation of affected populations [46–49].

Public policy management

Effective emergency evacuation planning during natural disasters depends on structured, data-driven coordination across agencies. Advanced Machine Learning (ML) techniques, including Deep Learning (DL) and Ant Colony Optimization (ACO), enable precise modeling of crowd dynamics, enhancing real-time decision-making in resource deployment and evacuation management [50–52]. Visualization tools that display escape routes, medical aid stations, and firefighting resources assist responders and evacuees in navigating high-stress environments efficiently [53], while AI-enabled fire detection systems deployed in vulnerable areas such as forests and commercial buildings offer early threat identification and faster response times [54, 55].

Post-disaster recovery requires a multidimensional assessment of urban resilience, where AI-powered data analytics provide insights into the economic, social, and cultural revitalization of affected communities, supporting strategic reconstruction planning [56]. In transport-related incidents, image recognition technologies optimize emergency responses by enhancing situational awareness and improving data precision, particularly in complex environments like railway systems [57]. AI frameworks also extend to managing violent mass casualty incidents—including active shooter scenarios—by optimizing evacuation routes and improving training simulations, thereby reducing casualties and enhancing operational preparedness [50].

In the public health domain, governmental intervention policies—such as curfews, travel restrictions, and age-specific limitations—have proven effective in reducing emergency department congestion and optimizing healthcare resource distribution during epidemics [33, 58] Beyond crisis response, AI-driven analytics and predictive modeling are increasingly applied in education policy, assisting educators in identifying students at risk of academic underperformance and enabling the design of targeted support strategies to improve outcomes [59, 60]. Collectively, these AI applications form a critical layer within broader disaster management systems, reinforcing both immediate response capabilities and long-term recovery planning.

Healthcare systems

Artificial Intelligence (AI) has emerged as a foundational technology in managing pandemics and health-related disasters by streamlining virus detection, diagnostics, drug development, and public health surveillance. Tools such as image recognition, deep learning algorithms, and facial recognition systems have optimized epidemiological tracking and healthcare resource allocation, significantly enhancing emergency preparedness during global crises like COVID-19 [61]. The pandemic accelerated the deployment of IoT-enabled humanitarian aid frameworks that support contactless suspect detection, remote patient monitoring, relief supply logistics, surveillance, and automated disinfection processes—collectively reducing transmission risks and improving the operational efficiency of healthcare systems [62].

AI-based systems also play a central role in clinical emergency management by predicting illness trajectories, assessing disease severity, and facilitating patient triage. Smart monitoring infrastructures, including biosensor-embedded wearables and automated clinical data processors, allow continuous real-time evaluation of patient vitals and deterioration risks, thereby advancing strategic interventions in high-pressure scenarios [63, 64]. Moreover, AI-enhanced diagnostic tools are particularly vital in identifying cardiovascular complications—conditions that disproportionately affect both disaster victims and first responders [65].

Beyond clinical settings, AI applications extend to aerial and satellite imaging for post-disaster assessment. Despite their value in documenting damage zones, conventional imaging systems often encounter delays in data processing and distribution. Synthetic Aperture Radar (SAR), with its capacity to generate high-resolution imagery under adverse weather and cloud cover, improves situational awareness and reliability in damage evaluation. Recent integration of Deep Learning (DL) algorithms into SAR analysis workflows has further automated interpretation processes, enabling more accurate damage mapping and more effective post-disaster recovery planning [50].

Information based approaches

Information-based approaches constitute a critical pillar of AI-enabled disaster governance by enabling proactive, data-driven responses across all phases of crisis management. Leveraging technologies such as Artificial Intelligence (AI), the Internet of Things (IoT), Geographic Information Systems (GIS), and big data platforms, these approaches enhance the accuracy, speed, and scalability of disaster preparedness, response, and recovery. They support dynamic risk assessment, early warning systems, real-time monitoring, and strategic resource allocation, thereby shifting crisis governance from reactive decision-making to anticipatory action. This section is organized into four subdomains: (1) analytical tools that process and interpret diverse data streams; (2) intelligent systems and digital platforms that integrate multiple technologies for operational efficiency; (3) economic and financial modeling for resource optimization and policy resilience; and (4) data governance frameworks that ensure security, interoperability, and informed decision-making. Together, these components enable equitable, efficient, and context-sensitive disaster response—particularly within health system infrastructures.

Analysis and tools

The integration of Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) has revolutionized disaster management by enhancing monitoring, prediction, and response capabilities. AI-powered early warning systems synthesize real-time environmental data, satellite imagery, and social media content to identify irregular patterns and initiate rapid intervention protocols [55]. Geographic Information Systems (GIS) complement these systems by enabling precise mapping of damage zones and resource allocation, supporting infrastructure recovery and urban planning [44, 66]. Moreover, Big Data Analytics and the Internet of Things (IoT) play a central role in urban disaster resilience, enabling continuous device communication, granular risk assessments, and advanced crisis forecasting [42]. IoT-based flood detection tools—such as accelerometers, borehole inclinometers, and pressure sensors—generate accurate impact metrics, empowering authorities to deploy timely protective measures for at-risk communities [2, 43].

Predictive ML models improve the precision of disaster forecasting and emergency logistics, allowing for optimized evacuation routes and targeted resource deployment [67]. AI-driven classification systems assess multi-hazard risks, including earthquakes, tsunamis, and tornadoes, enabling proactive mitigation strategies [43]. The combined use of GIS-based visual analytics and AI-enhanced decision support platforms bolsters situational awareness, facilitating strategic decisions regarding workforce deployment, medical equipment allocation, and epidemic control measures [55]. Additionally, wearable health technologies—such as smartwatches and RFID wristbands—integrated with AI systems, ensure the continuous monitoring of first responders’ vital signs, safeguarding their health during prolonged crisis operations [68–70].

Beyond environmental sensing, AI-augmented social media analytics extract actionable intelligence from real-time user-generated content, enabling dynamic threat identification and crisis assessment [46]. Integrated AI rescue optimization frameworks, which incorporate traffic monitoring, structural integrity evaluations, and GIS overlays, streamline emergency coordination efforts and accelerate operational efficiency [25]. Collectively, these analytical tools and intelligent systems reinforce disaster management ecosystems, promoting data-informed resilience planning and improving the scalability of future response initiatives [71, 72].

System and platform tools and applications

The growing dependence on digital technologies has led to the development of intelligent systems that significantly enhance the efficiency of emergency response and disaster management. AI-powered tools, such as decision support systems, assist emergency dispatchers in assessing incident severity and allocating appropriate medical resources using traffic-related datasets [73]. In high-risk industrial environments like underground mining, GUI-based command centers integrated with real-time sensor inputs play a vital role in guiding rescue and evacuation procedures [74]. The COVID-19 pandemic underscored the need for advanced Emergency Management of Public Health Events (EMPHE), where big data-driven platforms—such as the Big Data Administration and COVID-19 Epidemic Control Command—enabled strategic planning through knowledge distribution analysis and intervention trend identification [33, 75]. During the COVID-19 pandemic, machine learning algorithms, multi-objective optimization methods such as Pareto analysis, and graph neural networks enabled precise epidemiological data analysis, optimized treatment processes, and intelligent allocation of medical resources. Furthermore, digital contact tracing systems, powered by big data, proved effective in identifying sources of infection and predicting transmission pathways. These advancements highlight the immense potential of intelligent technologies to strengthen resilience, accelerate decision-making, and improve responsiveness in public health emergencies—a potential that can be extended to other areas of crisis management as well [76].

Novel systems combining real-time context recognition with emergency notifications have been developed to assess individual health risks, composed of components such as emergency, interaction, and context-aware managers [77]. Further studies have emphasized the importance of early detection systems in forecasting healthcare system strain [78]. In the context of environmental hazards, AI-driven predictive models are increasingly used, including neuro-fuzzy systems for urban flood forecasting based on rainfall patterns [79], machine learning algorithms to improve pandemic preparedness [80], and GNSS data applications for tsunami prediction [8]. Within the industrial sector, expert systems are implemented in refinery emergencies to ensure protocol optimization and personnel safety [81]. Tools like the Smart Medical Assistant for Disasters (SMAD), which incorporates deep learning-based sentiment analysis, provide integrated psychological and physical support to affected populations [82].

The convergence of Big Data Analytics (BDA) and the Internet of Things (IoT) has enabled real-time data acquisition during disasters, which informs policy decisions such as agricultural zoning in flood-prone areas [26, 83]. IoT tools like Node-RED, InfluxDB, and Grafana enhance situational awareness, particularly in fire emergencies, with future integration of machine learning expected to optimize these systems further [84]. Mobile applications like WISER, enhanced by machine learning, assist responders in identifying hazardous chemical agents based on symptoms [50, 85], while fire-specific decision support systems enable real-time crisis resolution [86]. Geographic Information Systems (GIS) have proven essential in planning road traffic emergency response across all disaster phases [87].

Social media analytics have significantly advanced disaster detection and monitoring. Platforms like Sina Weibo have outperformed traditional rule-based and machine learning methods in detection accuracy [46]. Tools such as Historian software log chronological crisis data to support post-event analysis [4], and Geocrawler extracts real-time disaster-related content from platforms like Twitter, YouTube, and Foursquare for continuous monitoring [88]. A novel approach combining Named Entity Recognition (NER) models with crowdsourced, geolocated Twitter content has enhanced the detection of disaster events [89]. Emergency Management Agencies (EMAs) now use smartphone-based social media tools to assess local crisis severity and coordinate effective responses [46].

Comprehensive disaster databases and platforms like the AMORE system contribute to evaluating the accessibility and equity of emergency services through visual and statistical tools [48]. Additionally, platforms using data mining, regression, and input-output modeling estimate economic losses from urban flooding, generating strategic defense reports [37]. Group decision-making during crises has benefited from dynamic Case-Based Reasoning Group Decision-Making (CBRGDM) models, which combine historical case retrieval with collaborative evaluation to formulate response alternatives [90]. In public health, AI-based early warning systems demonstrate greater predictive accuracy compared to traditional systems [91]. IoT-based platforms with system dynamics modeling further enhance risk detection, event scheduling, and emergency management in public health domains [32]. Finally, Real-Time Medical Emergency Response Systems, which process vast sensor data streams, allow for immediate and informed intervention during health crises [3].

Economic and financial analysis

Data-driven strategies have emerged as critical tools for evaluating and mitigating the economic consequences of public health emergencies and natural disasters. The integration of artificial intelligence (AI) in financial modeling enables real-time quantification of both direct and indirect economic losses, thereby enhancing scenario-based disaster planning and the efficiency of resource allocation [37, 92]. During the COVID-19 pandemic, the collaboration between MTN Nigeria and the Nigeria Governors’ Forum (NGF) exemplified the practical application of such approaches. By leveraging mobile money transaction data and telecommunications analytics, stakeholders were able to optimize the deployment of healthcare personnel and critical supplies to high-risk zones, significantly improving the timeliness and precision of emergency response efforts [93].

Simulation-based studies further indicate that data-driven infrastructure planning can substantially reduce energy consumption during post-disaster recovery phases, contributing to long-term economic sustainability and resilience [42]. In parallel, governments increasingly employ tax preference policies to stabilize financial markets and facilitate economic recovery in disaster-stricken regions. By modernizing tax administration systems with big data analytics, authorities can enhance epidemic surveillance, response capacity, and ensure broader macro-financial stability [92].

Moreover, the establishment of cross-provincial emergency logistics frameworks and centralized transportation hubs has proven essential in maintaining continuity of essential services in regions vulnerable to epidemic outbreaks. These logistics systems not only reduce economic disruptions but also bolster the resilience and adaptability of supply chains [92].

Data governance

Effective data governance is fundamental to ensuring reliable, scalable, and secure information management in contemporary disaster response systems. AI-driven models now process multimodal data—including text, images, and video—to support real-time decision-making and situational awareness [11, 25]. Multi-site data collection strategies improve information granularity and accuracy, reinforcing the integrity of crisis assessments. In large-scale applications, such as national disaster registries, consistency in syntax and semantics is crucial for efficient data retrieval and standardized analysis. In the healthcare domain, medical documentation must adhere to unified protocols to ensure interoperability, while IT infrastructure and institutional safeguards protect sensitive patient information [94].

Geographic Information Systems (GIS) enhance emergency responsiveness by delivering location-based intelligence, whereas Expert Systems (ES) integrate real-time field data and remote sensing to enable proactive disaster assessments and safety interventions [95]. Sensor-enabled data acquisition systems further support emergency preparedness, particularly in medical and traffic-related incidents, by linking live monitoring inputs with healthcare databases for early warning and rapid response [96].

Traditionally, impact assessments and post-disaster recovery estimates relied heavily on manual reports and insurance documentation. Today, AI-powered big data analytics offer automated, accurate, and scalable alternatives, streamlining the recovery planning process [35]. These innovations are especially critical in low- and middle-income countries (LMICs), where crises often stem from intersecting human and environmental vulnerabilities. In such contexts, AI-enhanced ICT tools generate localized insights, improving both resilience and operational efficiency. Future research should focus on tailoring data governance frameworks to specific disaster types, regional dynamics, and evolving situational demands [97].

The digital transformation of communication systems has redefined information dissemination during crises. While social media platforms like Twitter now serve as real-time disaster communication hubs, legacy systems often prioritize text-based data over multimedia formats. Updated data governance policies advocate for the integration of diverse content types, promoting dynamic interaction between the public and emergency services [68]. In parallel, scientific and governmental websites remain reliable sources of validated, though less frequent, information updates [4].

Specialized disaster datasets further support strategic planning and crisis management. For instance, Brazil’s national disaster registry compiles monthly data on disaster events, classifications, and socioeconomic variables such as income and the Municipal Human Development Index (IDH-M) [98]. Likewise, historical traffic accident data informs the design of real-time medical decision-support systems, enhancing the coordination of emergency medical services [73]. Real-time data capture should become a priority in future registries to increase responsiveness and operational precision [99]. Structuring unorganized disaster data is vital for computational modeling, particularly during health emergencies, where timely correlation and clustering algorithms can significantly improve forecasting and risk reduction [11].

A robust knowledge management framework underpins effective disaster data governance, encompassing data collection, processing, and long-term archival. Investments in digital storage capacity and AI-based analytics platforms strengthen institutional decision-making and public trust [4, 58]. Promoting data literacy and evaluating the functional readiness of disaster databases are equally important in maintaining system resilience. Without high-quality, accessible datasets, emergency response efforts risk becoming fragmented and inefficient. A comprehensive, well-implemented data governance strategy is essential for developing adaptable, transparent, and resilient disaster management infrastructures [58].

Process and operational methods

This category addresses the operational, logistical, and human-centered dimensions of disaster governance, focusing on how artificial intelligence (AI) facilitates coordinated actions across different sectors and communities. It encompasses four key pillars: social analysis and governance, social media analysis and communication strategies, supply chain management and logistics, and collaborative operational coordination. Together, these elements frame how AI can be employed not only to enhance situational awareness and optimize emergency response logistics, but also to foster inclusive and resilient governance structures through digital participation, real-time data analysis, and cross-institutional integration. By examining these operational mechanisms, this section highlights how technology-supported interventions can bridge gaps in preparedness, mitigate cascading failures, and promote equity and efficiency in disaster recovery efforts.

Social analysis and governance

Artificial Intelligence (AI) has become a critical component of modern disaster response frameworks, significantly enhancing crisis prediction, automated decision-making, and large-scale data analysis. In health-related emergencies, AI and machine learning (ML) models are widely utilized to forecast medical outcomes for high-risk populations, including displaced elderly individuals [100] and immigrants, whose vulnerability stems from unstable housing and limited access to healthcare services [101].

AI-driven approaches have expanded into mental health crisis management, particularly in university settings where psychological distress among students has prompted the implementation of big data surveillance and AI-assisted intervention programs [102]. ML techniques can also identify biological markers associated with post-traumatic stress disorder (PTSD), facilitating early detection and treatment planning shortly after traumatic events [103]. During public health crises, AI systems are deployed to detect and counter misinformation across digital platforms, improving the reliability of emergency communication and supporting real-time sentiment analysis and policymaking [36, 104]. Additionally, models for Network Public Opinion (NPO) analysis and emotion recognition provide authorities with actionable insights into public sentiment trends, helping refine crisis management strategies [105].

The COVID-19 pandemic exposed deep disparities in emotional well-being across populations. Research has shown that Americans experienced higher levels of psychological distress compared to Canadians and Mexicans [106]. AI-enabled ML models trained on hospital data have proven effective in predicting suicide risk within 90 days of emergency room visits, enabling preemptive mental health interventions [107]. As pandemic-related restrictions contributed to increased substance use and suicidality, the integration of AI-powered social resilience mechanisms has become essential to holistic disaster recovery planning [108].

Furthermore, lockdowns intensified mental health concerns, with AI studies identifying shifts in the prevalence of eating disorders, substance abuse, psychosis, and suicidality in emergency department records before and after the pandemic [109]. Strengthening disaster resilience in such contexts requires the application of AI-driven social media analytics, which facilitate public access to digital mental health resources and support adaptive recovery initiatives [5].

Social media analysis and governance

Social media platforms have become indispensable tools in emergency response, offering authorities real-time insights into public sentiment, misinformation trends, and crisis-related discourse. Platforms such as Twitter facilitate the extraction of crisis-specific information, enabling decision-makers to assess community reactions and prioritize urgent needs effectively [5]. Advanced semantic analysis technologies—such as Social Big Data (SBD) and Emergency Named Entity Recognition (ENER)—support the identification of both explicit and implicit emergency-related content, thereby enhancing the precision of response strategies and intervention efforts [10].

The pervasive use of social networks has significantly expanded the capacity to monitor public perception and emotional responses during disasters. Individuals share their thoughts, concerns, and emotions across diverse media formats, including text, video, and audio, allowing authorities to observe behavioral trends and shifts in public sentiment [45, 104]. Research suggests that influential opinion leaders play a critical role in combating misinformation by disseminating corrective content when it aligns with independently verifiable evidence and audience-specific communication patterns [47].

Social media has also emerged as a vital communication conduit during emergencies, providing the public with direct access to official updates from crisis management agencies while also enabling grassroots participation in relief operations [46]. Historians and digital researchers study these online interactions to derive behavioral insights that inform future crisis preparedness strategies. These findings can help governments and policymakers refine crisis communication protocols, thereby improving public trust, compliance, and community resilience [4].

Emotional analytics of social media content further illuminate the psychological impact of crises. For example, analyses of Weibo blog posts during disease outbreaks have revealed patterns in emotional progression, illustrating how continuous exposure to crisis narratives can influence public sentiment over time [49]. Concurrently, IoT-integrated platforms that monitor patient behavior, emerging health threats, and system anomalies in real time serve to reinforce early warning mechanisms and support proactive disaster mitigation [4].

Effective governance of social media ecosystems is critical to ensuring accurate and responsible dissemination of information. Studies examining media communication models during public emergencies highlight how uncontrolled information flow can increase social entropy, reinforcing the need for structured media oversight and engagement strategies [110]. Researchers advocate for improved journalistic standards, including enhanced content verification, higher-quality reporting, and the integration of AI-powered platforms to strengthen crisis communication. Closer collaboration between official and independent media outlets is also essential for fostering transparency and trust in emergency information delivery [110].

Supply chain management and logistics

Efficient supply chain management is a fundamental pillar in humanitarian disaster response, enabling authorities to forecast demand, allocate resources efficiently, and ensure timely distribution of critical supplies [98]. The World Health Organization (WHO) proposes a structured three-phase strategy for improving disaster logistics: (1) assessing vulnerability based on population demographics, particularly age distribution; (2) estimating disease transmission and case numbers using predictive modeling; and (3) determining essential resource needs in advance [93]. These steps help mitigate shortages, reduce delays, and ensure operational continuity during emergencies.

Technological advancements have revolutionized disaster logistics. The integration of Electronic Product Code (EPC) with Internet of Things (IoT) systems provides real-time visibility across the supply chain, especially useful during post-disaster municipal solid waste (MSW) management [111]. Automated systems equipped with Radio Frequency Identification (RFID) technology facilitate inventory control, track expired or missing supplies, and optimize warehouse organization, reducing inefficiencies and preventing stockouts or waste [112].

Simulation-based logistics models allow emergency managers to test scenarios, optimize distribution routes, and preempt bottlenecks [112]. Predictive analytics embedded in these models guide decision-making in dynamic crisis environments, supporting both speed and accuracy. Governments are encouraged to adopt pre-stocking protocols and centralized distribution planning to avoid redundancy and inventory imbalances [93, 98, 111, 112].

Recent developments in IoT-based emergency logistics systems, combining real-time monitoring, analytics, and automation, enhance situational awareness and streamline operations [112]. Workflow automation tools and digital financing frameworks have been introduced to improve fund allocation speed and transparency, strengthening long-term disaster resilience [112]. These innovations contribute to the creation of adaptive, data-driven supply networks capable of meeting the unpredictable demands of modern disasters.

Collaborative operational coordination

Ultimately, effective operational coordination depends on global collaboration, open data sharing, and trust-based governance. The development of open-access AI platforms that integrate epidemiological data would significantly enhance international crisis response capabilities [61, 64]. AI offers tools to monitor and predict population trends via search engine queries and social media analytics. These capabilities can facilitate early warnings for disease outbreaks and promote large-scale collaboration through digital platforms [4, 113].

Disaster and risk management require efficient interagency collaboration and data-driven coordination to ensure timely and effective responses. The Internet of Things (IoT) serves as a foundational infrastructure for early warning systems and rapid response strategies, with Community Sensing and Response (CSR) systems enabling real-time monitoring and informed decision-making [114]. While IoT enhances automated crisis detection, human-centered elements—such as public education, risk awareness, and coordinated emergency management—remain crucial to ensuring successful technology integration into disaster response frameworks. The effective deployment of IoT solutions also depends on international collaboration, standardized interoperability protocols, and comprehensive training for end users, all of which enhance disaster preparedness and operational resilience [114].

Big data analytics derived from IoT further bolsters disaster communication infrastructure by allowing authorities to anticipate network disruptions and optimize emergency information flow. By analyzing sensor data, traffic patterns, and user behavior, decision-makers can allocate resources more effectively and respond to shifting crisis dynamics [42]. In addition, geospatial analysis and population distribution data improve network planning for post-disaster search and rescue operations, ensuring that emergency responders receive timely and accurate information to guide strategic coordination and resource mobilization [42, 115].

Public health emergencies require an integrated communication framework that supports multi-agency coordination. AI-driven data analysis strengthens governmental responses, healthcare system management, and the distribution of medical resources [33]. For instance, AI-powered hospital dashboards can regulate patient inflows, manage logistics, and optimize emergency department operations to alleviate pressure on healthcare systems [33, 44, 66]. Moreover, AI technologies contribute to psychological crisis response through mobile-based Psychological Crisis Early Warning Systems, which incorporate mood tracking, mental health assessments, and remote consultation scheduling to promote timely intervention and reduce mental health stigma [102]. AI-driven training simulations also enhance preparedness by evaluating behavioral responses in high-stress scenarios, equipping emergency personnel with experience in real-world crisis conditions [50].

Global disaster coordination increasingly relies on AI-human hybrid models to enhance large-scale response strategies. The Artificial Intelligence for Disaster Response (AIDR) platform—developed in collaboration with the United Nations Office for the Coordination of Humanitarian Affairs (OCHA)—enables digital volunteers to label social media content during emergencies. These annotations are used to train machine learning models capable of detecting critical, time-sensitive information in real time [116]. Additionally, AI-based network optimization tools manage bandwidth allocation and maintain uninterrupted communication channels, ensuring secure and reliable information transmission in disaster zones [54]. Despite the advantages of AI, human oversight remains indispensable for ensuring ethical governance, protecting data privacy, and promoting equitable access to disaster response technologies [4, 61].

While ethical, legal, and technical challenges remain in applying AI to disaster governance, this review emphasizes practical use cases and highlights the need for further interdisciplinary research to address these emerging concerns comprehensively.

Conclusions

In the face of increasing global disasters, the integration of artificial intelligence (AI) into disaster governance presents a paradigm shift in how crises are managed, mitigated, and overcome. By systematically examining the potential of AI technologies, this paper highlights actionable insights that address both theoretical gaps and practical challenges in disaster preparedness, response, and recovery. Building on recent advancements in AI, the study offers a forward-looking perspective that aligns technological innovations with sustainable and equitable disaster management practices.

This study highlights the transformative role of artificial intelligence (AI) in disaster governance, presenting a conceptual framework that synthesizes existing knowledge and addresses key gaps in understanding and application. By categorizing AI applications into predictive analytics, decision-making, health system resilience, and social governance, the study provides actionable insights for advancing both theoretical understanding and practical implementation in disaster management.

Technologies such as Geographic Information Systems (GIS), machine learning models, and IoT-driven big data are shown to revolutionize disaster preparedness, response, and recovery by enabling real-time data integration, resource optimization, and enhanced situational awareness.

The research underscores the importance of leveraging AI to promote long-term resilience and equity in health governance systems. By systematically analyzing AI capabilities, the study charts a strategic roadmap for policymakers and researchers, fostering collaboration at national and international levels to tackle socioeconomic disparities and strengthen disaster resilience globally. Findings are categorized into three domains: governance functions and elements, information-based approaches, and process and operational methods. Governance functions emphasize enhancing government legitimacy, improving public policy mechanisms, and reinforcing healthcare system resilience. Information-based approaches focus on predictive analytics and real-time data integration to enable proactive disaster strategies, while process and operational methods address supply chain logistics, collaborative communication, and social media governance to facilitate coordinated responses and build societal resilience.

The study contributes to the field by unifying fragmented insights into a coherent framework, emphasizing innovative strategies that redefine disaster governance. Practical implications include fostering sound governance, strategic planning, and the adoption of AI technologies to mitigate disaster impacts, enhance societal resilience, and ensure equitable health outcomes. Future research should explore ethical, regulatory, and technical barriers to AI implementation and provide empirical studies to validate the conceptual framework presented.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

DL

Deep learning

AI

Artificial intelligence

ML

Machine learning

IoT

Internet of things

GIS

Geographic information system

Author contributions

“MR and PK conceived the study. MR wrote the main manuscript text and prepared figures. SA provided feedback on the result and manuscript. MR and AB categorized AI applications independently and created descriptions by synthesizing the extracted information. SA, PN, SMA, SR, and AB edited the manuscript. All authors reviewed the manuscript.”

Funding

This study was supported by the Research Center for Emergency and Disaster Resilience, Red Crescent Society of the Islamic Republic of Iran, Tehran, Iran.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study received ethical approval from the Research Ethics Committees of Education, Research, and Technology Division of Iranian Red Crescent Society (Approval ID: IR.RCS.REC.1404.001).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.