evidoria

← Back to browse

Good practice Imported

Seoul's AI 119 Call Bot — South Korea's First AI-Assisted Emergency Dispatch Triage

South Korea · Seoul · See the South Korea profile · See the Seoul profile

Evidence: Observational / pre–post Top 74% 47/100 · Ask Evidence Copilot about this practice

Seoul's Fire and Disaster Headquarters piloted South Korea's first AI call-taking system for its 119 emergency line in 2025, cutting call-handling time by about 9.6 seconds and triaging 11,434 reports (2,250 emergencies) in its first four months.

11,434 reports
Reports processed (first 4 months of pilot, 2025)
2,250 reports
Reports classified as emergencies (first 4 months of pilot, 2025)
9.6 seconds faster than voice-only intake
Call-handling time improvement (2025)
Seoul's AI 119 Call Bot — South Korea's First AI-Assisted Emergency Dispatch Triage

Details

Maturity
Pilot
Promoter
Seoul Metropolitan Government / Seoul Fire and Disaster Headquarters
Period
2025–present
Keywords
emergency services, public safety, 119/112 dispatch, disaster response

Context

Seoul's Fire and Disaster Headquarters launched South Korea's first AI-assisted emergency call-taking system for the 119 line in 2025, built to help secure the 'golden time' that determines survival in fires, cardiac arrests and other emergencies.

Objectives

The AI Call Bot is designed to listen to incoming reports, extract the location and type of incident, detect when multiple similar reports from the same area suggest a larger event such as a building collapse or spreading fire, and forward the most urgent cases to human dispatchers.

Activities

The system is designed to handle up to 240 simultaneous calls, with human operators at the Seoul Emergency Operations Center still making the final dispatch decision for urgent cases.

Results

In its first four months of pilot operation, the system processed 11,434 reports, of which 2,250 were classified as emergencies, and Seoul officials reported call-handling times about 9.6 seconds faster than voice-only intake.

Conclusions

These results are self-reported by the city government rather than independently audited, no public technical documentation on the model's accuracy or false-negative rate was found, and as of the sources reviewed no other Korean city had replicated the system.

Implementation
Indicative cost
Medium (€50k–€500k) — No public cost figures were found; conservatively banded medium for a city-scale AI call-triage pilot integrated with an existing emergency dispatch centre.
Time to results
Short (< 1 year) — Launched in 2025 and evaluated after its first four months of pilot operation; banded short given the brief reporting window.
Staffing & skills
Seoul Emergency Operations Center human dispatchers who make final dispatch decisions

Conditions for success

  • AI triage routes only the most urgent cases to human dispatchers, keeping humans in the final decision loop
  • Capacity to handle up to 240 simultaneous calls to manage surge events

Common failure modes

  • Results are self-reported by the city government with no independent audit
  • No public technical documentation on the model's accuracy or false-negative rate was found
  • No other Korean city had replicated the system as of the sources reviewed, so citywide generalisability is untested

Where it fits

Governance type
municipal government (Seoul Metropolitan Government / Fire and Disaster Headquarters)
Scale
single city, pilot phase
Income level
high-income

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

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.

Attachments

Similar practices you may find useful