AI Analyzes 911 Call Transcripts to Identify Violence Against Women 911 Calls: AI Helps Identify Violence Against Women Women
“911 – What’s your emergency?” The three-digit emergency hotline is known around the world. Operators dispatch police, firefighters, and ambulances to resolve dangerous situations and save lives. Fast and decisive action is needed to support victims of violence.
Analyzing Data to Detect Violence
But often, the immediate response to an incident is not followed by any measures to prevent future violence. This leaves victims in danger and their abusers feeling unpunished.
A new project by the United Nations Office on Drugs and Crime (UNODC) and Mexico’s National Institute of Statistics and Geography (INEGI) Mexico City uses artificial intelligence to sift through unstructured data and spot signs of violence: AI analyzes transcripts of 911 calls.
“AI is rapidly transforming sectors, and its application in public safety and the prevention of violence against women offers significant opportunities to create more inclusive and peaceful societies,” says UNODC Representative in Mexico Christian Holge.
Correctly classifying incidents
In most countries, emergency operators classify incoming calls based on whether they need help for burns, broken bones or other injuries. But this traditional approach can mean ignoring situations where an intimate partner is responsible for the injury.
Authorities are therefore missing opportunities to intervene. Victims of gender-based violence, often women at the hands of men, receive less support than they could. Child victims and families are also left behind.
A new UNODC and INEGI project aims to change that. It helps to identify and correctly classify cases of partner violence that were initially labeled as medical incidents or accidents.
More broadly, this pilot model demonstrates how AI applications can improve public services, including the police, social services, and the education system.
Privacy Protection
Data analysis is conducted according to strict legal standards to protect individuals’ privacy and ensure responsible handling of information.
A clear legal framework and data protection protocols will be key to scaling up the model more widely in Mexico and other countries.
The use of AI will help create a more comprehensive and integrated approach to public safety.
Beyond Mexico
In many countries, emergency response resources are limited and response centers are often overwhelmed. Misclassification, whether due to lack of training or the sheer volume of calls, leaves vulnerable populations unprotected.
The lessons learned from the UNODC and INEGI project could be replicated elsewhere. An AI-based approach could increase the capacity of public services to detect and respond to gender-based violence by improving classification mechanisms.
The initial results of the Mexico City project highlight the power of AI to drive positive social change.
“As this technology is used more widely, it could transform public safety and gender equality efforts around the world,” Holge said.