Breaking the language barrier: Bengaluru police launch India’s first AI-powered multilingual 112 helpline

In a city as linguistically diverse as Bengaluru, a distress call doesn’t always come in a language the listener understands. For the Bengaluru police, that gap has long posed a challenge — one that could mean the difference between swift response and unintentional delay.In a city as linguistically diverse as Bengaluru, a distress call doesn’t always come in a language the listener understands. For the Bengaluru police, that gap has long posed a challenge — one that can mean the difference between a swift response and a critical delay. Now, the force is turning to artificial intelligence to bridge it.On Wednesday, the Bengaluru police rolled out a first-of-its-kind multilingual upgrade to the Namma 112 emergency response system, designed to ensure that cries for help are no longer lost in translation. The AI-powered module is designed to handle calls in multiple languages, especially when callers speak tongues unfamiliar to control room staff. “Language should never become a barrier when someone is seeking help. With this initiative, we are making our emergency response system more inclusive and efficient. We are also exploring the addition of more languages, including Japanese, which is currently under discussion,” said Kuldeep Kumar Jain, Joint Commissioner of Police (Admin). From call to response: how Namma 112 works Bengaluru’s emergency response system, Namma 112, is currently staffed by around 200 personnel, including call operators, supervisors and police staff, working across three shifts. Each shift has approximately 50 staff members, while the night shift operates with a reduced strength of around 35 to 40 personnel, including supervisors. Speaking to The Indian Express, Jain said the control room receives between 9,000 and 10,000 calls a day, of which nearly 2,000 are actionable.Story continues below this ad Until now, operators primarily handled calls in English, Kannada, Hindi, Tamil, and Telugu. But Bengaluru’s cosmopolitan mix — shaped by migrants, professionals and tourists — means emergency calls often come in a far wider range of languages. In such cases, operators would switch to English or Hindi to gather basic details, a workaround that sometimes slowed response times. The new system, developed under the Multilingual Voice AI for Nationwide Key Interventions (VANKI) initiative in collaboration with Monday Ventures and Aeos, aims to make these interactions quicker and more precise. Highlighting the recent upgrade, Home Minister G Parameshwara said the city’s emergency response system now supports communication in more than 10 languages. “Bengaluru is an internationally renowned city, and people from across the globe visit us. Language was a major barrier for those seeking help from the police, especially through the control room. Now, we are the first police force in the country to interact in more than 10 languages,” he said at a press conference Wednesday.Story continues below this ad “This will greatly assist those who are not familiar with the local language and require help during emergencies. This is a first-of-its-kind initiative in India, and we hope other states will follow,” he added. Inside the multilingual response system When a call comes in, a human operator answers it first. If there is a language mismatch, the call is transferred to the AI system. The AI identifies the language, interacts with the caller, and asks key questions to gather details such as location and the nature of the emergency. It then generates a Call for Service (CFS) request in real time, while a human operator monitors the exchange to ensure accuracy. The upgrade follows an internal survey of callers’ linguistic diversity. In addition to major Indian languages, the system now supports Malayalam, Marathi, Gujarati, Odia, Kashmiri, Manipuri and Bengali, along with foreign languages such as French, Arabic, Spanish and Nepali. More languages are expected to be added in phases. Officers say the system will help reduce delays