[REVISION HISTORY]

India AI integration in public governance

Updated 2 times since CLSTR started tracking revisions of this situation.

What changed

2026-08-13 16:14 UTC → 2026-08-16 13:01 UTC · added ~~removed~~

India is advancing the integration of artificial intelligence into its public administration and digital infrastructure. Initial efforts focused on cybersecurity, with the Indian Computer Emergency Response Team (CERT-In) deploying AI models to audit approximately 500 government applications for vulnerabilities. Through the Digital India programme, the government is scaling platforms like UMANG and DigiLocker. As of late July 2026, UMANG provides approximately 2,575 services to over 11.66 crore registered users, while DigiLocker serves more than 72.43 crore users with 5,437 document types. To improve inclusivity, the Digital India BHASHINI Division and NITI Aayog are developing the ‘BHASHINI for Seva/Sanchalan’ initiative. This partnership integrates language translation APIs and voice-first AI to allow citizens to access services in their preferred Indian languages. Digital access is further supported by over 407,000 functional Common Service Centres at the Gram Panchayat level, alongside tools like SMRITI for multilingual transcription and NITI TARA for grassroots data-driven decision-making. At the state level, governments are adopting specialized AI for social welfare and administration. Kerala has partnered with SAS to use advanced analytics for improving beneficiary selection and disaster response. Meanwhile, Maharashtra has approved the deployment of Sarvam AI’s Indus platform for approximately 2,500 officials. This sovereign AI workspace assists with research, drafting, and translation across 22 languages, including Marathi, using secure, audited infrastructure located within India. As of August 2026, the government is expanding these multilingual efforts. The BHASHINI platform now hosts over 360 AI-based language models providing machine translation and speech recognition. Furthermore, the BharatGen project, anchored at IIT Bombay, is developing a multimodal Large Language Model designed to support all 22 scheduled Indian languages.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.