Democratising AI in India

1. At a Glance

• Democratising AI = expanding affordable access to AI compute, datasets, models, skills and applications across India, anchored in the IndiaAI Mission under MeitY [1][3].
• Vision: "Making AI in India and Making AI Work for India" — building sovereign, inclusive, India-centric AI [2].
• Examinable as a GS-III (S&T / economy) and GS-II (governance) flagship; converges with Digital India, 5G rollout, DPI stack [1][3].

2. Why in the News

• PIB Backgrounder "Democratising AI in India" released 10 February 2026 highlighting 38,000+ GPUs at ₹65/hour, AIKosh's 7,500+ datasets and 273 models, and 99.9% district 5G coverage [1].
• Two-year progress report of IndiaAI Mission (May 2026) consolidating achievements [3].

3. Background & Evolution

• 2018: NITI Aayog's National Strategy for AI ("AI for All") — first policy framing [2].
• 2020: National AI Portal (INDIAai.gov.in) launched by MeitY + NASSCOM [2].
• March 2024: Union Cabinet approved IndiaAI Mission with outlay of ₹10,371.92 crore over 5 years [2].
• March 2025: AIKosh platform launched by MeitY as a unified repository of datasets, models, sandbox [4].
• June 2025: BharatGen — first govt-funded, indigenous multimodal LLM supporting 22 Indian languages — launched [5].
• Feb 2026: PIB Backgrounder consolidates progress [1].

4. Core Static Facts

• Nodal Ministry: Ministry of Electronics & Information Technology (MeitY) [2].
• Implementing arm: IndiaAI, an Independent Business Division (IBD) under Digital India Corporation [2].
• Outlay: ₹10,371.92 crore / 5 years (approved March 2024) [2].

• Seven Pillars of IndiaAI Mission [2]: 1. IndiaAI Compute Capacity 2. IndiaAI Innovation Centre (IAIC) 3. IndiaAI Datasets Platform (AIKosh) 4. IndiaAI Application Development Initiative 5. IndiaAI FutureSkills 6. IndiaAI Startup Financing 7. Safe & Trusted AI

• Compute: 38,000+ GPUs empanelled (initial target was 10,000) at ₹65/hour subsidised rate [1][2].

• AIKosh: 7,541 datasets + 273 AI models across 20 sectors [1][5].
• 5G coverage: 99.9% of districts — telecom backbone for AI [1].
• BharatGen: indigenous multimodal LLM, 22 Indian languages, launched 2 June 2025 [5].
• IndiaAI Safety Institute: hub-and-spoke model; 13 Responsible AI projects (machine unlearning, bias mitigation, privacy tools) [5].

5. Multi-Dimensional Analysis

Economic / Technological

• Subsidised GPU access (₹65/hr) reduces compute cost barrier for startups, MSMEs, academia [1].
• Indigenous foundation models (BharatGen, Sarvam) reduce dependence on foreign LLMs [5].
• IndiaAI Startup Financing pillar deploys risk capital into deep-tech AI ventures [2].

Social / Inclusion

• Multilingual coverage (22 scheduled languages) addresses linguistic equity beyond English [5].
• AIKosh = "national digital public good" → equal access for Tier-2/3 institutions [1].
• FutureSkills Prime extends AI skilling beyond metros [2].

Governance / Ethical

• Safe & Trusted AI pillar + IndiaAI Safety Institute mandate Responsible AI: bias mitigation, machine unlearning, privacy-enhancing tech [5].
• Aligns with Digital Personal Data Protection Act, 2023 obligations.

Strategic / Geopolitical

• Sovereign compute + sovereign models reduce strategic dependence amid US export controls on advanced chips [2].
• India chaired GPAI (Global Partnership on AI) and hosted GPAI Summit Dec 2023; positions India in global AI governance [2].

Administrative

• Public–Private Partnership (PPP) model for GPU empanelment; State/UT partnerships for AI CoEs [5].
• Hub-and-spoke architecture for Safety Institute pulls in academia, industry, startups [5].

6. Recent Developments (12–18 months)

• March 2024: Cabinet nod ₹10,372 cr to IndiaAI Mission [2].
• March 2025: AIKosh launched [4].
• June 2025: BharatGen unveiled at BharatGen Summit [5].
• Feb 2026: 38,000 GPUs milestone announced; PIB Backgrounder issued [1].
• 2-year progress report released highlighting startup financing, 13 Responsible AI projects [3][5].
• AI Centres of Excellence being established across States/UTs in collaborative model [5].

7. Prelims Hooks

• IndiaAI Mission outlay: ₹10,371.92 crore over 5 years [2].
• Approved by Union Cabinet in March 2024 [2].
• Nodal ministry: MeitY (not DST, not NITI Aayog) [2].
• Implementing arm: IndiaAI IBD under Digital India Corporation [2].
• IndiaAI Mission has 7 pillars [2].
• AIKosh = datasets + models platform; 7,541 datasets, 273 models, 20 sectors [1].
• 38,000+ GPUs at ₹65/hour [1].
• 5G: covers 99.9% of districts [1].
• BharatGen launched 2 June 2025, supports 22 Indian languages, multimodal [5].
• National Strategy for AI ("AI for All") released by NITI Aayog in 2018 [2].
• IndiaAI Safety Institute runs 13 Responsible AI projects [5].
• India hosted GPAI Summit 2023 in New Delhi [2].

8. Mains Relevance

• GS-III: Science & Technology — developments, indigenisation; Economy — employment, IT/ITeS.
• GS-II: Governance — e-governance, DPI; international institutions (GPAI).
• GS-IV: Ethics in technology — AI bias, accountability.
• Sample stems: 1. "Discuss how the IndiaAI Mission seeks to democratise access to AI while balancing safety and sovereignty." (GS-III, 250 words) 2. "AI is the new electricity. Examine India's strategy to make AI a public good." (GS-III) 3. "Algorithmic bias and opacity raise ethical dilemmas in AI governance. Comment in the Indian context." (GS-IV)

9. Related Topics to Study Next

• Digital Public Infrastructure (DPI) — AI rides on Aadhaar/UPI/DigiLocker rails.
• National Strategy for AI, 2018 (NITI Aayog) — policy parent document.
• Semicon India Programme — chip ecosystem feeds GPU/compute sovereignty.
• Digital Personal Data Protection Act, 2023 — data law underpinning AI datasets.
• GPAI (Global Partnership on AI) — international AI governance forum.
• 5G & BharatNet — connectivity layer for AI deployment.
• Skill India / FutureSkills Prime — AI talent pipeline.
• National Quantum Mission — adjacent frontier-tech mission.

10. Common Errors / Trap Areas

• IndiaAI Mission is under MeitY, not NITI Aayog (NITI authored the 2018 strategy, not the mission) [2].
• AIKosh ≠ INDIAai portal; AIKosh is the datasets/models repository launched 2025 [4].
• IndiaAI is an IBD under Digital India Corporation, not a statutory body.
• BharatGen is govt-funded indigenous LLM; do not confuse with Sarvam (private; also on AIKosh) [5].
• Budget is ₹10,371.92 cr / 5 years, not annual.
• GPAI Summit was hosted by India in 2023 (Delhi), not 2024.