- The Government of India white paper emphasises that AI progress relies on computing power, datasets, and model ecosystems, positioning AI infrastructure as a strategic public asset for inclusion, competitiveness, and sovereignty.
India’s Current AI Infrastructure Gap
- Data–Capacity Mismatch: India generates ~20% of global data but hosts only ~3% of global data centre capacity, forcing reliance on foreign compute infrastructure.
- Compute Access Bottleneck: India’s common/national compute capacity has crossed ~34,000 GPUs, but demand from startups, academia and government far outstrips supply.
- Frontier Entry Barrier: Globally, advanced chips & hyperscale compute are controlled by a few players, raising entry costs and limiting India’s ability to train frontier models.
AI Ecosystem in India at Present
- Tech Growth: India’s technology sector projected to earn USD 280 billion in 2026.
- AI Workforce: Over 6 million people employed in India’s tech and AI ecosystem.
- Global Centres: 1,800+ Global Capability Centres, including 500+ focused on AI.
- Startup Ecosystem: 1.8 lakh startups; 89% of new startups use AI solutions.
- Sector Impact: Industrial, retail, BFSI, and healthcare sectors contribute 60% of AI value.
Core Idea of the White Paper
- Infrastructure Over Algorithms: AI leadership depends on who controls compute, data and models, not only who builds apps or chatbots.
- Foundational Asset: Compute power, datasets, and AI ecosystems are emerging as core economic assets like roads/electricity for a modern economy.
- National Priority: For India, AI infrastructure is linked to competitiveness, inclusion and digital sovereignty, not only technology.
|
How India Plans to Build AI Infrastructure?
India is building robust AI infrastructure through national compute expansion and trust-centric, inclusive digital platforms to ensure wider access and governance.
Capacity Build-Up
- IndiaAI Mission Push: IndiaAI is building a national AI compute infrastructure with 18,000+ GPUs through PPP models, enabling wider access across sectors.
- AI Supercomputing Backbone: AI supercomputer AIRAWAT (C-DAC Pune) was ranked 75th globally (Top500 list), strengthening the public AI compute backbone.
DPI Approach
- Language Inclusion: Bhashini supports multilingual AI tools and citizen-facing services, helping reduce language barriers in Digital India delivery.
- Trust-Centric Governance: DPI-based AI access emphasises standards, interoperability and accountability. E.g., DigiLocker enables consent-based, verifiable document sharing across government services.
Critical Risks in AI Infrastructure
- Compute Concentration: Training a frontier AI model can cost $10 million–$100 million+, so only a few firms/countries can afford large-scale compute access.
- Uneven Sector Adoption: AI adoption is concentrated in finance/e-commerce/IT, while high-impact sectors like agriculture/health/education lag, deepening productivity gaps.
- Sustainability Constraint: AI infra expansion is power and water-intensive; India’s data-centre growth is already flagged for resource stress in major hubs.
- Weak Accountability: If AI access expands without audit trails, the risk of model misuse, like deepfakes, rises sharply with low enforcement capacity.
Way Forward
- Sovereign Capacity: Strengthen domestic compute and strategic tech capability to reduce dependency on external clouds and chips; E.g., National Supercomputing Mission for public AI research.
- Green Infrastructure: Mandate energy-efficient architecture, advanced cooling and renewable linkage for large AI/data centres; E.g., Singapore’s green data-centre efficiency norms.
- Sector Inclusion: Push AI adoption in lagging sectors using shared infrastructure for the public good. E.g. ABDM enabling AI-driven diagnostics, and AgriStack enabling precision advisory for farmers.
- Compute Commons: Create affordable shared GPU access for startups, universities and state governments through pooled national capacity.
“Making AI in India and making AI work for India” can drive inclusive societal development, ensuring technological autonomy and global competitiveness; as Sundar Pichai notes, “AI is one of humanity’s most important tools for progress.”
Reference: The Hindu
PMF IAS Pathfinder for Mains – Question 519
Q. To what extent is access to AI infrastructure increasingly viewed as a matter of digital sovereignty for India? Examine the challenges India faces and suggest policy measures to democratise AI infrastructure. (250 Words) (15 Marks)
Approach
- Introduction: Write a brief introduction about India’s AI ecosystem by mentioning the recent data.
- Body: Write how AI infrastructure is increasingly viewed as a matter of digital sovereignty for India, mention challenges, and suggest policy measures to democratise AI infrastructure.
- Conclusion: Emphasis on Indigenising AI and its infrastructure is key to democratise access, drive innovation, and strengthen India’s digital sovereignty.
