New Orleans Deploys Carbyne AI to Triage 911 Emergency Calls

Summary

The Orleans Parish Communication District (OPCD) in New Orleans has begun implementing Carbyne's AI Emergency Call Triage to assist with emergency 911 dispatch. The AI system analyzes incoming calls and identifies whether they are related to an ongoing incident. If the call is related to an existing incident, the AI provides information or updates to the caller; otherwise, it routes the caller to a human dispatcher. This deployment follows the successful implementation of Carbyne's AI on 311 non-emergency calls in April, where it handled 50% of calls requesting information.

New Orleans receives over 1,000 emergency calls per day, making the dispatch system one of the busiest in the United States. The AI triage system is designed to reduce the volume of calls human dispatchers must handle by filtering out duplicate calls related to the same incident during periods of high call volume. Importantly, Carbyne's AI is not intended to handle emergency calls directly—it serves as a routing and information tool to prioritize calls and connect callers to appropriate resources or human dispatchers.

The implementation does carry potential risks. The article notes concerns about AI reliability, including the possibility of hidden biases in algorithm decision-making and potential extensions to predictive policing applications. Additionally, automatic speech recognition systems may struggle with callers who have strong accents, dialects, or different speech patterns. Experts emphasize that effective AI 911 triage requires robust oversight, thorough training data, strong cybersecurity, and proper algorithm regulation to ensure the system operates safely and equitably.

• Deployment carries risks including speech recognition failures with accented speech and potential algorithmic biases that could reinforce systemic inequities