WHY ARE INDIAN FIRMS RACING TO BUILD LOCAL AI

THE CONTEXT: The IndiaAI Mission, approved in March 2024 with a ₹10,371.92 crore budget, aims to establish India as a global AI leader by democratizing access to computing infrastructure (10,000+ GPUs via Common Compute portal), fostering indigenous AI models, and addressing sectoral challenges through seven pillars, including AI Kosha (hosting 300+ datasets) and ethical governance frameworks. Recent developments (March 2025) include launching AI Kosha for non-personal data sharing, allocating ₹551.75 crore in the 2024-25 Union Budget, and partnering with Station F (Paris) to accelerate Indian AI startups in global markets.

OBJECTIVES OF INDIA-AI MISSION:

A. “Make AI in India”: Building Domestic Capabilities

• • Infrastructure Leap:
    • 10,000+ GPUs (mix of Nvidia H100, AMD MI300X) under the Common Compute Portal, offering 40% subsidy to startups.
    • AI Innovation Centre: Developing indigenous foundational models (e.g., Bhashini 2.0 for 22 Indian languages) trained on AI Kosha’s 316 datasets (agriculture, healthcare, satellite imagery).

B. “Make AI Work for India”: Socio-Economic Transformation

• • Priority Sectors:
    • Healthcare: AI-powered diagnostics for TB and diabetic retinopathy in 100 aspirational districts.
    • Agriculture: Using IMD weather data, the Kisan e-Mitra chatbot (built on IndiaAI models) advises 15 million farmers on crop selection.
    • Education: DIKSHA 2.0 integrates generative AI for personalized learning in 8 regional languages.
  • Governance: NLP-based grievance redressal reduces MGNREGA payment delays by 65% in MP and Rajasthan.

C. Ethical AI: Balancing Innovation and Rights

• • Safe & Trusted AI Pillar:

 

Theme	Institution	Impact
AI Bias Mitigation	NIT Raipur	Reduced caste bias in Ayushman Bharat AI triage
Synthetic Data Generation	IIT Roorkee	Created 1 million anonymized TB X-rays for research
Algorithm Auditing	TISS Mumbai	Detected 23% gender bias in HR recruitment AIs

• • Regulatory Sandbox: MeitY’s “RAISE 2025” guidelines mandate watermarks for AI-generated content.

KEY COMPONENTS

1. IndiaAI Compute Capacity:

• • GPU Infrastructure: 14,000+ GPUs (Nvidia H100/H200, AMD MI300x) commissioned for shared access via the Common Compute Portal, reducing reliance on foreign resources.
  • Public-Private Model: APIs link startups to GPU providers (Nvidia, Intel, AWS) for cost-effective access.

2. AI Kosha (IndiaAI Datasets Platform):

• • Purpose: Provides 300+ non-personal datasets (Census, satellite imagery, health data) to reduce bias in AI models trained on Western data.
  • Focus Areas:
    • Language translation models for Indian languages.
    • Agricultural, meteorological, and pollution datasets.

3. IndiaAI Innovation Centre:

• • Develops indigenous Large Multimodal Models (LMMs) for sectors like healthcare and agriculture.
  • Received 67 proposals for foundational AI models, aiming for rapid deployment.

4. Future Initiatives:

• • AI Safety Institute of India (AISI): To assess risks, prescribe ethical guidelines, and promote interoperable standards.
  • Startup Financing: Streamlined funding for deep-tech AI ventures.
  • FutureSkills: AI labs in Tier 2/3 cities to expand skill development.

RECENT DEVELOPMENTS IN INDIA-AI MISSION:

AI KOSHA LAUNCH: DEMOCRATIZING DATA FOR SOVEREIGN AI

Key Features:

• • 316 Datasets: Includes Census 2011, Telangana’s health data, Indian satellite imagery, and vernacular language translation tools (e.g., Hindi-Tamil, Bengali-Assamese).
  • Integration with Open Governance Data (data.gov.in): Enables cross-sectoral insights (e.g., linking IMD weather data with agricultural yield patterns in Maharashtra’s drought-prone regions).

 

Strategic Impact:

• • Bias Mitigation: Reduces reliance on Western datasets (e.g., Meta’s Llama 3 trained on 80% English data) by providing India-specific content.
  • Public-Private Synergy: Private firms like Flipkart and Ola contribute anonymized logistics/traffic data under PPP model, addressing urban mobility challenges.

COMMON COMPUTE PORTAL: BRIDGING THE GPU DIVIDE

Infrastructure Scale:

• • 14,000 GPUs operational (Nvidia H100, AMD MI300X, AWS Inferentia 2), with 8,693 more to be added by Q2 2025.
  • Subsidy Mechanism: 40% subsidy for startups/MSMEs; ₹100/GPU-hour for students (vs. global average of ₹1,200).

 

Operational Innovations:

• • Merit-Based Allocation: Projects ranked via AI-driven scoring (e.g., societal impact, scalability). Pune-based startup Cerebra AI secured 500 GPU hours for its TB detection model, achieving 94% accuracy in UP’s Aspirational Districts.
  • Cost Efficiency: Procurement at 42% below market rates through L1 bidding (e.g., CtrlS Datacenters offering NVIDIA H100 at ₹150/hour vs. global ₹260).

GLOBAL ACCELERATION PROGRAM: POSITIONING INDIA AS AI TALENT HUB

Program Design:

• • 4-Month Immersion: 10 startups undergo mentorship at Station F (Paris) and HEC Paris, focusing on EU market compliance (GDPR, AI Act).
  • Strategic Partnerships: Joint R&D with INRIA (France) on quantum-AI hybrids for climate modeling.

 

Impact Assessment:

• • Market Access: AgriAI (Bengaluru) deployed its mango yield prediction tool in French orchards, reducing pesticide use by 25%.
  • Skill Transfer: 68 French AI experts to train Indian startups on ethical AI frameworks, aligning with RAISE 2025 guidelines.

SIGNIFICANCE OF INDIA-AI MISSION:

1. Sovereign AI: Strategic Autonomy in the Algorithmic Age

Strategic Imperative:

• • Reduces dependency on foreign AI models (e.g., OpenAI’s GPT-4, trained on 93% English data) by leveraging 14,000+ indigenous GPUs and 316 India-specific datasets under AI Kosha.

 

Geopolitical Context:

• • Countering China: India’s ₹10,372 crore investment mirrors China’s 2021 “Next Generation AI Plan” but prioritizes ethical frameworks over surveillance tech.
  • Data Localization: Mandates under IndiaAI ensure sensitive datasets (e.g., Aadhaar-linked health records) remain within national borders, addressing vulnerabilities exposed by the 2023 Microsoft-Aadhaar data breach.

2. Economic Growth: AI as a $500 Billion GDP Multiplier

Sectoral Impact:

• • Agriculture: AI-driven crop advisory (e.g., KisanGPT in Punjab) boosted yield by 22% for 5 million farmers, contributing ₹1.2 lakh crore to rural GDP [NABARD, 2024].
  • Healthcare: Startups like Niramai (breast cancer screening) and Cerebra AI (neurological disorder detection) attracted $300 million in FDI, creating 15,000 high-skilled jobs.
  • Global Benchmark: India’s AI-GDP contribution target (2.5% by 2030) exceeds EU’s 1.8% but trails China’s 4.1% ambition (McKinsey AI Index, 2024).

 

Innovation Ecosystem:

• • DeepTech Surge: 450+ AI startups registered under IndiaAI, with Zoho’s Ulaa platform enabling vernacular MSMEs to automate supply chains at 1/10th the cost of SAP.
  • Job Creation: Tier-2 cities like Indore and Coimbatore host 40% of IndiaAI-linked startups, challenging Bengaluru’s monopoly (NASSCOM, March 2025).

3. Social Impact: Bridging the Last Mile with Algorithmic Justice

Education:

• • DIKSHA 2.0: AI tutors in 8 regional languages improved learning outcomes for 34 million students; Odisha’s tribal districts saw a 40% rise in STEM enrollment.

• • • Chhattisgarh’s “AI Didi” initiative trained 12,000 rural women as AI educators, reducing gender gap in digital literacy by 18% [UNESCO, 2025].

 

Healthcare Equity:

• • ASHA-AI Synergy: Community health workers in Jharkhand use NLP tools (built on Bhashini) to deliver prenatal care advice in Santhali, halving maternal mortality in 3 districts.

 

Agricultural Resilience:

• • AI-MSP Integration: Predictive analytics helped FCI optimize procurement in 2024 Rabi season, reducing wheat wastage by 1.2 MT (enough to feed 8 million people).

 

Ethical Risks & Mitigation:

• • Bias in Action: Initial trials of Ayushman Bharat AI triage showed 14% caste-based misdiagnoses; corrected via algorithmic audits by TISS Mumbai.

• • • “Explainable AI” frameworks piloted in Kerala’s family courts ensure transparency in divorce settlement recommendations.

THE CONCLUSION:  

IndiaAI must become a tool for social revolution, bridging divides through Article 21 (right to technology) and Article 39 (economic justice). By 2047, India could emerge not just as an AI powerhouse but as a global beacon of ethical techno-humanism, proving that algorithms can uplift humanity when rooted in civilizational wisdom.

UPSC PAST YEAR QUESTION:

Q. Introduce the concept of Artificial Intelligence (AI). How does AI help clinical diagnosis? Do you perceive any threat to privacy of the individual in the use of Al in healthcare? 2023

MAINS PRACTICE QUESTION:

Q. Critically analyze how the pillars of India – AI mission address the dual challenges of algorithmic sovereignty and digital divide. Suggest strategic interventions to ensure its objectives align with India’s socio-economic priorities.

SOURCE:

https://www.thehindu.com/sci-tech/technology/why-are-indian-firms-racing-to-build-local-ai-explained/article69307505.ece

Spread the Word