AI Emergency Alerts Are Coming. Humans Might Not Be Ready
wired ·
As climate disasters become more frequent, emergency managers are increasingly turning to generative AI to improve public communication during crises. While human officials often struggle with jargon and clarity, leading to potentially fatal misunderstandings like those in Montecito, California, AI tools can draft precise, actionable alerts in seconds. Companies such as EM1 have developed models that outperform many human experts in synthesizing best practices for disaster messaging. Despite this potential, official adoption remains low due to concerns about public trust and the perception that AI-generated content lacks authenticity. Many emergency managers are already using chatbots unofficially, though they fear backlash from skeptical populations or increased warning fatigue. Experts warn that without proper oversight, inexperienced users may fail to evaluate AI outputs correctly, risking further misinformation. The transition to automated alerts represents a critical test of whether technology can enhance safety without eroding the public's confidence in government agencies.
wired