evidoria

← Back to browse

Good practice Imported

NADMO AI-Powered Multi-Hazard Chatbot — UNESCO Helps Ghana's Disaster Agency Reach Citizens Before the Water Does

Ghana · Accra · See the Ghana profile · See the Accra profile

Evidence: Descriptive / self-reported Top 75% 47/100 · Ask Evidence Copilot about this practice

UNESCO handed Ghana's National Disaster Management Organization an AI-powered chatbot that answers citizen questions on floods, fires and storms in multiple languages around the clock; officers in ten regions were trained to run it, though no usage data has been published yet.

10 regions
NADMO regional offices with chatbot operators trained online (as of 4 June 2026)
6 regions
Additional regions scheduled for in-person chatbot training

Details

Maturity
Scaling
Promoter
National Disaster Management Organization (NADMO), Ghana, with UNESCO and funding from the Government of Japan
Period
Handed over May 2026
Keywords
disaster management, emergency communication, public safety

Context

During floods and other emergencies, Ghanaian disaster officials have long relied on radio bulletins and phone lines that congest exactly when demand is highest, leaving many citizens without timely, individualised information.

Objectives

Deliver automated advisories on floods, fires, storms, droughts, earthquakes and epidemics on demand and during active emergencies, in multiple languages so warnings are not limited to English-language broadcasts alone.

Activities

On 26 May 2026, UNESCO handed Ghana's National Disaster Management Organization (NADMO) an AI-powered multi-hazard chatbot, developed under the project "Strengthening resilience to water-related disasters under climate change for a sustainable society in Ghana" and funded by the Government of Japan. On 4 June 2026, officers from ten regional NADMO offices (Greater Accra, Ashanti, Eastern, Central, Western, Volta, Northern, Bono East, Upper East and Upper West) were trained online on both the citizen-facing interface and the back-end system, with six further regions scheduled for in-person training.

Conclusions

As a system handed over only weeks before this review, no public usage, reach or impact data exists yet — the deployment is real and multi-region, but its effectiveness as a disaster-communication tool during an actual emergency has not yet been evaluated or reported.

Implementation
Indicative cost
Low (< €50k)
Time to results
Short (< 1 year)
Staffing & skills
NADMO regional officers trained on the citizen-facing interface and back-end configuration system, UNESCO technical development team, NADMO national coordination staff overseeing rollout across regions

Conditions for success

  • Completion of in-person training in the remaining six regions
  • Ongoing multi-language content configuration and maintenance on the back end
  • Citizen awareness of and access to the chatbot channel during an actual emergency
  • Publication of usage/impact data once the system is tested in a real disaster event

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful