Current Affairs Topics Archive
International Relations Economics Polity & Governance Environment & Ecology Science & Technology Internal Security Geography Social Issues Art & Culture Modern History

Experts Explain | India AI applications stack: What the country can bring to the table

February 13, 2026 5 min read Science & Technology Polity & Governance GS3 (Science & Technology Digital Infrastructure Economic Development)
On This Page ▼

What Happened

  • India's AI ecosystem is increasingly positioned to offer a distinct "applications stack" to the world — leveraging its unique strengths in digital public infrastructure (DPI), a large pool of English-speaking technical talent, and a massive domestic data-generating population.
  • The IndiaAI Mission, approved by Cabinet on March 7, 2024, has emerged as the central government framework for building compute infrastructure, foundational AI models, datasets, and sector-specific applications.
  • By 2025, India had deployed over 38,000 GPUs under the IndiaAI Compute Pillar, in partnership with NVIDIA; this target was set to climb to 50,000 GPUs by end of 2025.
  • BharatGen — India's first government-funded multimodal large language model (LLM) initiative — secured ₹988.6 crore under the IndiaAI Mission, with aims to develop models up to one trillion parameters covering Indian languages, speech, and computer vision.
  • India ranks first globally in AI skill penetration according to the Stanford AI Index 2024, with a score of 2.8, ahead of the United States (2.2) and Germany (1.9).

Static Topic Bridges

IndiaAI Mission — Structure and Seven Pillars

The IndiaAI Mission was approved by the Union Cabinet on March 7, 2024, with a total outlay of ₹10,370 crore. The mission is structured around seven pillars that together constitute India's sovereign AI strategy: (1) Compute Capacity — building GPU-based AI cloud infrastructure; (2) Innovation Centre — establishing applied AI research centres; (3) Datasets Platform — providing access to high-quality, non-personal datasets for researchers and startups; (4) Application Development Initiative — supporting sector-specific AI applications; (5) FutureSkills — building AI skills and literacy at scale; (6) Startup Financing — providing risk capital for AI startups; and (7) Safe & Trusted AI — ensuring ethical, responsible AI development. The Union Budget 2025-26 allocated ₹2,000 crore to the mission — approximately one-fifth of its total outlay.

  • Mission approved: March 7, 2024; nodal ministry: Ministry of Electronics and Information Technology (MeitY).
  • Total outlay: ₹10,370 crore; Budget 2025-26 allocation: ₹2,000 crore.
  • 38,000+ GPUs deployed under compute pillar by 2025; target of 50,000 GPUs set for end of 2025.
  • 12 indigenous foundation models selected from 500+ proposals are being developed under the mission.
  • BharatGen received ₹988.6 crore as the largest single beneficiary under the IndiaAI Mission 2025.

Connection to this news: The IndiaAI Mission is the institutional backbone of India's AI applications stack — providing the compute, data, and model development infrastructure on which sector-specific applications across government and private sector are being built.

Digital Public Infrastructure (DPI) — India Stack as AI Foundation

India's Digital Public Infrastructure (DPI) — commonly called India Stack — is a set of open, interoperable digital platforms built on three foundational layers: identity (Aadhaar, launched 2009), payments (UPI, launched 2016), and data sharing (Account Aggregator, DigiLocker). India Stack has enabled financial inclusion, direct benefit transfers, and digital commerce at scale. As AI adoption accelerates, DPI provides unique advantages: massive structured datasets (transaction records, health data, agricultural data), real-time data generation at scale, and a common interface for AI-powered service delivery in government. The IndiaAI Dataset Platform aims to harness this DPI data for AI training, providing anonymised, high-quality datasets to researchers and startups. BHASHINI, an AI-led language translation platform, is designed to extend digital services to Indian language users — including voice-based access.

  • Aadhaar: 1.4 billion enrolled; foundation of identity verification across government schemes.
  • UPI processed over 100 billion transactions in 2023-24; generates massive real-time financial behaviour data.
  • Account Aggregator (AA) framework allows consented data sharing between financial institutions — a regulated data pipeline for AI.
  • BHASHINI supports over 22 Indian languages; designed to enable voice-based access to the internet in regional languages.
  • India's DPI model has attracted global interest; G20 India Presidency (2023) made DPI a flagship agenda item.

Connection to this news: India's AI applications stack is not built from scratch — it leverages a decade of DPI investments. The argument for India's unique AI contribution is precisely this DPI foundation: large, structured, government-grade datasets and interoperable platforms that few other countries possess at India's scale.

BharatGen and Sovereign AI — India's Foundational Model Strategy

BharatGen is India's sovereign AI initiative focused on developing indigenous Large Language Models (LLMs) and multimodal AI models with capabilities specifically tuned for Indian languages and contexts. The initiative was launched by Union Minister Jitendra Singh in 2024 at the BharatGen Summit. BharatGen's flagship model, Param-1, is a bilingual LLM with 2.9 billion parameters, pretrained on 5 trillion tokens in English and Hindi. More ambitiously, BharatGen is developing a 17-billion-parameter Mixture-of-Experts (MoE) model using the NVIDIA NeMo framework. The broader IndiaAI foundational model programme supports 12 indigenous foundation models across language, vision, and science domains. Digital India BHASHINI, a related initiative, provides AI-powered translation services across 22 scheduled languages, including voice-based access.

  • BharatGen described as the world's first government-funded multimodal LLM initiative.
  • Param-1: 2.9 billion parameters, pretrained on 5 trillion tokens in English and Hindi.
  • BharatGen's 17-billion-parameter MoE model developed using NVIDIA NeMo framework.
  • Total BharatGen funding: ₹988.6 crore from IndiaAI Mission.
  • Stanford AI Index 2024: India ranks first globally in AI skill penetration (score: 2.8 vs US: 2.2).

Connection to this news: BharatGen represents India's attempt to develop sovereign AI models — reducing dependence on US or Chinese AI models for government functions and Hindi/Indian language applications, which is the unique element India brings to the global AI applications stack.

Key Facts & Data

  • IndiaAI Mission approved: March 7, 2024; total outlay ₹10,370 crore.
  • Budget 2025-26 allocation for IndiaAI Mission: ₹2,000 crore.
  • GPUs deployed under compute pillar: 38,000+ (target 50,000 by end 2025).
  • BharatGen funding: ₹988.6 crore; largest single beneficiary of IndiaAI Mission.
  • Param-1 LLM: 2.9 billion parameters, pretrained on 5 trillion English+Hindi tokens.
  • India's AI skill penetration rank: 1st globally (Stanford AI Index 2024), score 2.8.
  • BHASHINI supports 22+ Indian languages including voice-based access.
  • 12 indigenous foundation models being developed from 500+ proposals received under the mission.
  • UPI: 100+ billion transactions in 2023-24 — foundational data layer for AI applications.