Soon, callers to Bengaluru’s 112 emergency helpline will be able to seek help in Spanish, French, Arabic, or Indian languages like Bengali, Manipuri, or Kashmiri, thanks to an AI-based multilingual voice agent that is being introduced by the city police.
Aimed at breaking language barriers that hinder access to emergency services — particularly for migrant workers, tourists, students —the AI system will be integrated into the Emergency Response Support System (ERSS 112).
Also read: AI Cops! Bengaluru police pilots AI-generated avatars for social media outreach
Bengaluru city police have already floated a tender for its installation and maintenance. “The AI voice assistant can understand and communicate in multiple languages. Initially, it will support Indian languages such as Bengali, Gujarati, Odia, Manipuri, Kashmiri, Nepali, and Assamese, in addition to foreign ones including Spanish, French, and Arabic. It can be quickly trained to include more languages as needed,” reads the tender document.
There are around 100 staff members in 112 helpline, with Kannada, English, and Hindi as primary languages. Each shift has about 30 staff and some are also able to manage Tamil, Telugu, and Malayalam. Officials said additional languages can be supported using AI tools.
It may be recalled that Bengaluru Police recently denied reports that a helpline operator asked a Spanish tourist to speak in Kannada during a burglary emergency call. They later clarified that no calls were made at the time of the incident, and action was taken the next morning based on a neighbour's detailed report.
Helpful in emergency situations
Unlike traditional helplines, AI-based multilingual voice agents can adapt to different emergency scenarios. For example, in a missing person report, it changes its questions based on whether the missing individual is a child, adult, or elderly person. It can also interact with emergency dispatch systems to coordinate ambulances, fire brigades, and disaster response teams, while location-tracking features help identify callers’ locations more accurately.
The system is designed to handle thousands of calls at once, which is particularly useful during disasters or large-scale emergencies when human operators are overwhelmed. “In such situations, every second counts. Having an AI that can instantly respond in someone’s native language could be the difference between life and death,” a senior official said.
“Bengaluru attracts people from across India, but not all speak Kannada or English—the languages used in emergency call centres,” said the officer. “This AI system ensures language isn’t a barrier to urgent help.”
Also read: Karnataka: Govt departments need AI cell nod for artificial intelligence projects
The new system also tackles the issue of staff shortages. “Recruiting multilingual emergency operators is both difficult and expensive. Training takes time, and even trained personnel can falter under pressure. In contrast, AI voice agent works round-the-clock without fatigue, delivering consistent and accurate responses”, the officer said.
It’s also cost-effective. Unlike traditional call centres that require physical infrastructure and large teams, the AI operates digitally and scales easily with call volume. This helps reduce long-term costs for the government. “Importantly, the system is not entirely autonomous. In complex or emotionally sensitive cases, the AI can quickly route the call to a human operator. This hybrid model ensures the efficiency of AI is balanced with human empathy and judgment” he said.
“Language should never be a barrier to getting help,” said a police official. “With this system, a caller from Assam or even Spain—can reach 112 and be understood. It’s a major step towards universal, accessible emergency services,” he added.
Also read: Bengaluru Police building AI-driven digital 'Punchnama' with body-worn cameras
Discover the latest Business News, Sensex, and Nifty updates. Obtain Personal Finance insights, tax queries, and expert opinions on Moneycontrol or download the Moneycontrol App to stay updated!
