Indigenous AI Development
India is rapidly advancing its artificial intelligence capabilities through the IndiaAI mission. The government is focused on making AI accessible and accountable for real-world applications. This initiative aligns with the vision of Prime Minister Narendra Modi to democratise technology and improve sectors such as healthcare, agriculture, education, climate, and governance.
Development of Foundational AI Models
IndiaAI supports the creation of indigenous Large Language Models (LLMs) and smaller language models trained on Indian datasets. These models understand local dialects and cultural nuances better than foreign models. Four startups—Sarvam AI, Soket AI, Gnani AI, and Gan AI—are developing these foundational models. The models will be open source to encourage startups to build India-specific AI applications.
Compute Infrastructure and Access
The mission has provisioned over 34,000 GPUs through the IndiaAI Compute Portal. These GPUs include Nvidia H100 and H200, Intel Gaudi, and AMD MI300X, among others. The portal provides discounted GPU access at about ₹67 per GPU-hour, roughly one-third of the global average. This infrastructure supports academia, startups, MSMEs, research bodies, and government agencies.
IndiaAI Datasets Platform (AIKosh)
AIKosh offers over 1,000 curated datasets across sectors such as healthcare, agriculture, and education. Examples include farmer queries from Kisan Call Centres, geological data, and clinical imaging for brain lesion diagnosis. The platform also hosts 208 AI models, including text-to-speech models in Indian languages like Bengali and Kannada, promoting inclusive AI development with strong data privacy safeguards.
AI Applications and Startup Support
IndiaAI funds around 30 AI applications addressing public interest problems in governance, health, and climate. Startups receive support to develop ethical and responsible AI solutions. The IndiaAI Startups Global programme, in partnership with Station F and HEC Paris, helps ten Indian AI startups expand into Europe. These startups focus on areas such as privacy-enhancing AI and cybersecurity.
Safe and Trusted AI Framework
The IndiaAI Safety Institute coordinates efforts on AI safety, security, and trust. It engages academia, industry, startups, and government in a hub-and-spoke model. Projects focus on responsible AI themes like bias mitigation, privacy tools, AI governance testing, watermarking, and deepfake detection. India will host the AI Impact Summit in February 2026, continuing its leadership in global AI governance.
Legal and Regulatory Measures
India employs a techno-legal approach combining laws and technology to manage AI risks. The Information Technology Act, 2000, addresses identity theft, misinformation, and deepfakes. The Bharatiya Nyay Sanhita, 2023, strengthens cybercrime provisions. The Digital Personal Data Protection Act, 2023, enforces data fiduciary accountability and citizen data rights. The IT Rules, 2021, impose obligations on intermediaries to remove harmful AI-generated content promptly. A dedicated cybercrime portal and toll-free number enhance citizen grievance redressal.
Techno-Legal Innovation and Research
India funds R&D at premier institutions like IITs to develop AI tools for privacy protection, deepfake detection, and cybersecurity. This ensures AI governance is backed by technological innovation rather than relying solely on legislation. The model balances innovation with safety and accountability, encouraging responsible AI growth.
| Related | |
|---|---|
