New Orleans Tests Carbyne's AI to Triage 911 Emergency Calls, Stresses It Won't Replace Human Operators
New Orleans' emergency call center is using Carbyne's AI to filter duplicate calls about the same incident, easing the burden on staff, while confirming that serious emergencies are still routed to human call-takers as before
The Orleans Parish Communication District (OPCD), the agency overseeing the emergency call center for New Orleans, Louisiana, is deploying an "AI Emergency Call Triage" system from Carbyne to help manage the volume of incoming emergency calls. The news became a topic of debate on Hacker News, earning 72 points and more than 117 comments.
At the core of the system is the filtering of duplicate calls about the same incident. One example cited is a road accident that generates a flood of simultaneous calls from multiple witnesses — a classic problem that forces dispatchers to waste time taking the same report over and over while new emergencies may be waiting in the queue.
Here's how it works: when a call comes in, the system first checks whether it relates to an incident that has already been reported. If so, the system automatically provides the caller with updated information. But if the call is unrelated, or involves another type of emergency such as a violent crime or a medical incident, it is immediately transferred to a human operator.
The point the agency emphasizes is that the system is not replacing call-takers, but rather serving as a supplementary tool for triage and call-volume management during major incidents, allowing staff to focus more on emergencies requiring urgent assistance. The OPCD issued this clarification to local media outlet WWL-TV/Nola.com after news reports framed the story as AI taking over 911 calls from humans.
Nevertheless, bringing AI into emergency systems remains a closely watched issue, since triage errors could cost lives. And with no real-world performance figures released yet — such as the proportion of calls the AI handles successfully or its error rate — the system's true effectiveness will have to be proven through long-term use.
This is a real-world example of AI being applied to public safety infrastructure. The policies and lessons learned here could serve as a model for other emergency call centers around the world, including Thailand's 191/1669 systems, which face similar call-overload problems during major incidents.