UPSC Mains Answer Practice — GS 3: India’s adoption of artificial intelligence and primary healthcare (Week 172)

Question 1: India's growing adoption of artificial intelligence has not been matched by an equivalent capacity to develop frontier AI technologies. This question is relevant to GS Paper III under Science and Technology, specifically focusing on emerging technologies and indigenous technological capabilities. It evaluates India's ability to transition from AI adoption to frontier AI development.

This understanding is essential for appreciating the challenges that India faces in maintaining global technological competitiveness and achieving strategic autonomy.

The answer should begin by acknowledging the multifaceted nature of artificial intelligence and its wide-ranging applications, from productivity and employment to scientific research, cybersecurity, and defence. India has established a robust digital infrastructure and is actively incorporating AI applications. However, the capacity to create frontier AI models is limited by insufficient computing capacity, infrastructure, and regulatory support.

Technological constraints include the reliance on large-scale computing capacity for frontier AI development. India's IndiaAI Mission has 45,000 GPUs, yet the 4,096 GPUs allocated to Sarvam AI for model training are significantly lower than those used by leading frontier models. This disparity creates a structural disadvantage that cannot be overcome solely through software innovation.

Additionally, India's potential to develop frontier models is contingent on access to cutting-edge computing infrastructure and powerful electronic equipment. The absence of adequate domestic capability in the broader technological hardware ecosystem increases reliance on overseas suppliers.

Another significant challenge is the limited availability of computing resources within India. While large data centers are being constructed, these facilities primarily serve global firms rather than Indian researchers and AI companies. This situation indicates that the geographical presence of data centers in India does not necessarily ensure the availability of computing resources to Indian entities.

India's startup ecosystem is flourishing in AI application development; however, there remains a significant gap between building applications on existing foundation models and inventing foundational or frontier models. Frontier AI holds crucial implications in scientific research, cybersecurity, and defence, making it imperative to distinguish between AI adoption and frontier AI development.

Policy constraints include the frequent emphasis on AI applications over competing in the frontier. While an application-focused approach can yield immediate economic benefits, excessive reliance on foreign foundation models may render India technically dependent. The national compute pool is insufficient for ongoing frontier-model development, necessitating the adoption of public policy measures to treat compute capability as a critical technical infrastructure rather than merely a commercial resource.

Environmental impacts associated with large data centers must also be considered. Their substantial energy and resource requirements may impose financial burdens on local communities. Consequently, enhancing India's compute capacity must be accompanied by appropriate environmental safeguards rather than being pursued at the expense of sustainability.

The policy dilemma extends beyond the creation of more powerful models. Artificial intelligence also enables sophisticated attacks, fraud, misinformation, and manipulation. The Finance Minister recently acknowledged AI as a "double-edged sword," highlighting the importance of balancing technical innovation with responsible governance and security protections.

In conclusion, India cannot simply be an efficient adopter of AI developed overseas. Frontier AI will increasingly impact economic competitiveness, scientific discovery, cybersecurity, and strategic power. Therefore, India requires a dual strategy that leverages AI extensively across the economy while simultaneously investing in sovereign compute, hardware, research, and frontier-model capabilities.

Points to Ponder: The distinction between AI adoption and frontier AI development lies in the scope and ambition of the technology. AI adoption involves utilizing existing foundation models for practical applications, whereas frontier AI development entails creating advanced, innovative models with significant implications in various domains. High costs, infrastructure constraints, and limited private investment hinder indigenous AI research by creating barriers to resource acquisition and technological advancement.