As artificial intelligence (AI) technologies are increasingly integrated into emergency management, ethical considerations demand greater attention. Essential components of comprehensive emergency management include mitigation, preparedness, response, and recovery, which should serve as the foundation for integrating AI-driven science and technologies to effectively safeguard populations and infrastructure in times of crisis. This paper reviewed the ethical challenges of AI in emergency management in terms of critical issues, best practices, applications, emerging ethical considerations, and strategies addressing ethical challenges. Three core ethical themes are identified: algorithmic bias; privacy, transparency and accountability; and human–AI collaboration. This paper thoroughly analyzed the associated ethical challenges, reviewed the theoretical frameworks and proposed strategies to mitigate ethical challenges by strengthening the audits of algorithms, enhancing transparency in AI decision-making, and incorporating stakeholder engagement. Finally, the importance of creating policies to govern AI ethics was discussed.

Disaster preparedness is one of the most important efforts undertaken in disaster management. It involves not only emergency management practitioners but also the whole community. Yet many adults in the US are not prepared. Using a user-centered designed VR disaster game, the potential to test, analyze, and operationalize variables that influence preparedness is evaluated based on results from experiments among older adults. This paper describes the process of developing game-based learning to improve individual and household disaster preparedness and to broaden the scope and analysis employed in disaster preparedness research and practice. Testing the following research questions, the paper relies on results from expert assessment surveys and in-person immersive experiments: (1) Are there decision-making variations among older adults in response to different hazards? (2) Are there differences in protective action among older adults?, and (3) How could emergency managers use data from a VR game on household disaster preparedness? The results indicate that the game could capture real-time simulated data on household preparedness using disaster theories and could identify areas of concern across multiple hazards. While older adults were generally eager and able to use VR, they require training. Older adults’ interest expanded to wanting more information about preparedness. The game captured six of the top ten data needs among the emergency management community. The implications to transform how preparedness research is collected and improve what practitioners know about individual preparedness. However, a dashboard and user-friendly interface are required; this paper proposes what one might contain.