India pushes for voice-based AI models, moving beyond text


By MYBRANDBOOK


India pushes for voice-based AI models, moving beyond text

The demand for India-specific voice AI models is growing, as global AI systems struggle with the country's linguistic diversity, with limited training data available for Indian languages despite supporting multiple languages

 

The government is taking significant steps to develop home-grown voice-based AI models, aiming to address the needs of a larger population that prefers vocal interaction with technology over text. This initiative is a shift from the global dominance of text-driven AI models like OpenAI’s ChatGPT, Google’s Gemini, and Elon Musk’s Grok, which primarily support English and a few other languages. India’s focus is on building AI systems optimized for speech and capable of supporting the nation’s linguistic diversity.

Abhishek Singh, Additional Secretary at the Ministry of Electronics and Information Technology (MeitY), explained that many people in India find it easier to communicate in their native language through voice rather than typing. “The Internet will become more voice-enabled, and there will be many people who will prefer accessing services using voice commands,” Singh said. Singh leads the IndiaAI mission, a Rs 10,372 crore initiative aimed at advancing India’s AI capabilities.

India has approximately 900 million active internet users, but nearly 500 million people are still offline, many of whom are from rural areas or have limited access to digital tools. For these communities, voice-based AI could be transformative, as it allows users to interact with technology in their local dialects. For example, a farmer could ask an AI assistant for information on crop diseases or weather updates in their native language, without the need to navigate a text-based interface. Similarly, voice-based systems could be used to access health advice, government schemes, or educational content.

Democratizing AI with voice technology

The demand for India-specific voice AI models is pressing, as international AI systems struggle to support the country’s rich linguistic diversity. While global models may support multiple languages, their training data for Indian languages remains limited. Singh emphasized that AI models developed with Indian datasets would provide more effective solutions and outperform existing global models.

To drive this initiative forward, the government has received 67 proposals from startups, academic institutions, and private companies. These proposals focus on developing large language models and smaller, domain-specific models for sectors like agriculture, healthcare, and education. A technical committee is currently evaluating these proposals, with selected projects set to receive government funding.

Additionally, the government is working on a national dataset platform that will provide anonymized data for training AI models, supporting the development of voice-based AI solutions. Through this initiative, the government aims to democratize AI, bridge the digital divide, and position India as a leader in voice-based AI innovation.

 E-Magazine 
 VIDEOS  Placeholder image

Copyright www.mybrandbook.co.in @1999-2025 - All rights reserved.
Reproduction in whole or in part in any form or medium without express written permission of Kalinga Digital Media Pvt. Ltd. is prohibited.
Other Initiatives : www.varindia.com | www.spoindia.org
of images belongs to the respective copyright holders