Research on Failed Emergency Responses

What issues have effected previous natural disaster responses, and how they might effect our idea.

https://www.adjustersinternational.com/resources/news-and-events/why-many-disaster-plans-fail-and-what-you-can-do-to-avoid-it/

Not practising results in problems only being found during the emergency.
Poor communication to people ranging from CEOs to children, leading to poor understanding of what they must do.
Too little information means important details are missing. Too much information means important details are clumped together and difficult to find.
Following disaster measures designed for vague scenarios rather than your specific location can cause issues when those specifics become relevant.

People have ignored natural disaster experts and emergencies due to mistrust. Experts have made mistakes with under developed tech or over-confidence, as well as conspiracies. Case in point, Italy’s “Power dictates, ‘science’ obeys, justice absolves” rally.

In Turkey, the organisation capable of responding the best to earthquakes is the military (should our A.I. have to communicate with militaries as well?).
The military were not given a significant role for Turkey’s emergency response, meaning they were too late to respond to an earthquake in February 2023.
Poor coordination and allocation of roles also contributed to a messy response (our A.I. is being planned to predict and call attention to disasters, not organise an emergency response).
One major effect to the response was the government’s insistence on suppressing information rather than utilising it (do we want to account for political interference for our A.I.?)