# US city introduces AI to help manage 911 emergency calls > Source: > Published: 2026-08-08 12:20:10+00:00 New AI technology that will handle emergency 911 calls is set to be implemented in New Orleans to reduce the volume of calls humans have to handle. In April, The Orleans Parish Communication District (OPCD) implemented an AI to answer 311 non-emergency calls. As 50% of 311 calls are callers seeking information, the AI was trained to provide just that, according to [GovTech.](https://www.govtech.com/artificial-intelligence/new-orleans-is-latest-to-answer-nonemergency-calls-with-ai) However, the OPCD is currently testing out a new tool intended to reduce the volume of emergency calls handled by human dispatchers, which sits at roughly 1,000 a day. ## New Orleans has one of the highest emergency call rates in the U.S. The OCPD uses Carbyne’s AI Emergency Call Triage. The system doesn’t handle emergency calls directly, but analyzes and prioritizes them, and provides immediate feedback to callers. It is intended to handle the number of calls that are related to one incident. Callers are automatically routed to an AI agent who asks if they are calling regarding the specific incident. If they are, they can receive information or updates. If they aren’t, they are transferred to a human dispatcher. While the AI system can be a useful tool regarding response times and dispatch efficiency, there may be some negative downsides to the system. If not properly monitored, the AI could develop hidden biases based on the data it is trained on. A [report by Police1](https://www.police1.com/artificial-intelligence/can-ai-fix-911s-biggest-problems-or-make-them-worse) suggests that the AI may have its predictive policing algorithms skewed by historical crime data from over-policed areas or marginalized communities. Essentially, it could reinforce socioeconomic bias and systemic racism. ## Related Equally, if the AI’s automatic speech recognition system fails to understand callers with stronger accents, dialects, or differing voice pitches, it could become unreliable in its responses. For more on AI, check out our coverage of [the viral AI chatbot that turned out to be one man trolling thousands](https://www.dexerto.com/entertainment/viral-new-ai-chatbot-is-actually-just-one-man-trolling-thousands-of-users-3396046/), or [Google Earth’s AI tool that was withdrawn after a day](https://www.dexerto.com/entertainment/google-earth-withdraws-ai-tool-after-users-fake-natural-disasters-3393966/).