AI Emergency Alerts Are Coming. Humans Might Not Be Ready

Emergency managers at state and local levels are increasingly turning to generative artificial intelligence to draft disaster alerts and public safety communications, raising both operational hopes and concerns regarding public trust.

The Communication Challenge in Disasters

State and local emergency managers form the front line of defense during extreme weather and other crises. However, human emergency personnel frequently struggle with public communication under pressure. Experts note that humans tend to use confusing jargon, omit vital context, or include irrelevant details. Past communication failures during wildfires and mudslides have demonstrated that unclear warnings can lead to tragic confusion or warning fatigue.

The Potential of AI in Warnings

Artificial intelligence systems can rapidly ingest extensive research on effective messaging, process thousands of prior alerts, and draft clear warnings within seconds. Proponents argue that an AI model can outperform a stressed human official struggling to recall training while alarms ring during an active crisis.

• Drafting Assistance: Officials like Ingham County emergency director Rob Dale have used AI behind the scenes for practice drafts.
• Software Solutions: Startups such as EM1 have trained models on disaster communication research, producing test alerts that outperform typical human drafts by clearly stating locations, instructions, and agency details.

Adoption Realities and Risks

While official agency adoption remains relatively low, experts report that unofficial usage of consumer chatbots like ChatGPT and Gemini is widespread among understaffed departments.

• Trust Concerns: Critics warn that AI-generated content may feel inauthentic, potentially damaging public trust in government agencies.
• Quality Control: Inexperienced users might fail to catch inaccuracies, as demonstrated by early incidents where officials generated flawed maps or unverified text.

Experts emphasize that while AI can accelerate drafting, human oversight remains critical to verify accuracy and maintain public confidence.