AI as Digital Public Infrastructure: India’s Vision for Affordable AI

Subject: GS Paper 3 : Indian Economy

Context: India is exploring the idea of making Artificial Intelligence (AI) the next Digital Public Infrastructure (DPI) by making AI inference affordable, interoperable, and widely accessible.

About Digital Public Infrastructure (DPI)

• Digital Public Infrastructure (DPI) refers to a set of foundational, interoperable digital systems built on open standards and protocols that provide essential public services to enable inclusive, efficient, and secure delivery of government and private-sector services at scale.
• Key Characteristics
  • Open and interoperable digital architecture.
  • Public-good approach with open standards and APIs.
  • Inclusive and scalable, ensuring universal access.
  • Enables innovation and competition by allowing public and private entities to build value-added services.

UPSC Online Classes

India’s Digital Public Infrastructure (DPI) Model

• Aadhaar: Universal digital identity platform, covering over 1.4 billion people, transforming identity verification into a low-cost API-based service.
• Unified Payments Interface (UPI): Interoperable real-time digital payments platform, processing around 20 billion transactions per month at near-zero cost.
• DEPA/Account Aggregator: Consent-based data-sharing framework enabling secure and user-controlled exchange of financial and personal data.

Distinctive Features of India’s DPI

• Integrated digital stack: India is the first country to build an integrated DPI combining identity (Aadhaar), payments (UPI), and data exchange (DEPA/AA) at population scale.
• Open and interoperable architecture: Enables seamless interaction between government agencies, financial institutions and private innovators through open APIs.
• Public digital rails: Government builds the foundational infrastructure, while private entities innovate and compete on service delivery.
• Inclusive and scalable: Designed for universal access, low transaction costs and mass adoption.

Impact of India’s DPI 

• Affordable internet access: The cost of 1 GB of mobile data declined from about US$4 in September 2016 to below US$0.30 by 2019, making India one of the world’s cheapest mobile data markets.
• Mass digital inclusion: Nearly 500 million people came online within five years, accelerating digital adoption.
• Global leadership: While countries such as Estonia (digital identity), Brazil (Pix payments), Singapore (Singpass and SGFinDex), and the European Union (open banking and data portability) have developed individual components, India is unique in integrating all three foundational pillars into a unified Digital Public Infrastructure (DPI) ecosystem.

Rationale for AI as the Next Digital Public Infrastructure (DPI)

• Foundational Digital Utility: Artificial Intelligence (AI) is emerging as a core digital utility, similar to identity, payments, and data, requiring public digital infrastructure for universal access.
• Builds on India’s DPI Success: Extends the Aadhaar–Unified Payments Interface (UPI)–Data Empowerment and Protection Architecture (DEPA) model by making AI affordable, interoperable, and widely accessible through open digital rails.

• Democratise Affordable AI Access: Focuses on reducing the cost of AI inference (using trained models) to enable widespread adoption by citizens, startups, businesses, and public institutions.
• Strengthen Technological Sovereignty: Reduces dependence on proprietary Artificial Intelligence (AI) models and Application Programming Interfaces (APIs) owned by foreign technology companies, while mitigating risks from pricing changes, export controls, and policy shifts.
• Leverage India’s Comparative Advantages: Utilises India’s AI talent, multilingual datasets, and large digital user base to build indigenous AI capabilities and foster innovation.

Click to Know UPSC Offline Courses

Proposed Framework for AI as the Next Digital Public Infrastructure (DPI)

• Affordable Compute Infrastructure: 
  • Expand Compute Capacity: Scale up Graphics Processing Unit (GPU) availability under the IndiaAI Mission through a public-private partnership.
  • Affordable Compute Access: Provide startups and researchers access to subsidised GPU compute (around ₹65 per GPU hour).
  • Reliable Energy Infrastructure: Integrate AI compute demand into the National Electricity Plan and ensure dedicated renewable and nuclear power for AI infrastructure.
• Open-Source Foundation Models:
  • Develop Indigenous AI Models: Build open-source Indic Large Language Models (LLMs).
  • Leverage Public Datasets: Make anonymised public datasets (e.g., legal rulings, agricultural data, educational curricula in all 22 scheduled languages) available for AI research.
  • Promote Open AI: Require AI models developed using publicly funded compute or datasets to be released under open-weight licences.
• Unified Intelligence Interface (UII): 
  • Create an AI Gateway: Develop a Unified Intelligence Interface (UII), similar to the Unified Payments Interface (UPI), to provide interoperable access to AI services.
  • Standardise AI Services: Establish common standards for identity, consent, billing, and safety across AI applications.
  • Enable Affordable AI Access: Facilitate low-cost AI access for startups, educational institutions, and government agencies.
• Token-Based AI Access
  • Freemium model: Provide monthly free AI API tokens to Aadhaar-verified startups and students.
  • Cross-subsidisation: Commercial users pay for higher AI usage, subsidising the free tier.
  • Support research and education: Divert one-tenth of the fertiliser subsidy to provide AI tokens for research institutions and schools.
  • Promote inclusive AI adoption: Expand access to AI across education, healthcare, agriculture and public service delivery.

Check Out UPSC CSE Books

Visit PW Store