01 Read
What happened
India's sovereign AI strategy is moving beyond access to global foundation models toward building homegrown AI applications suited to Indian languages, agriculture, healthcare, and governance needs. Policymakers and technologists argue that adopting foreign-built models without contextual adaptation fails India's diverse population. The push now focuses on fine-tuning models on Indian datasets, developing inference infrastructure, and deploying AI in public services — reflecting a shift from capability acquisition to real-world deployment for Bharat.
02 Understand
Why it matters
India's AI policy operates under the broader IndiaAI Mission, approved in March 2024 with a ₹10,372 crore outlay. The mission has several pillars: compute infrastructure (IndiaAI Compute Capacity), a curated dataset platform (IndiaAI Datasets Platform), a foundational model development push, and application deployment across sectors.
The distinction between 'foundational models' and 'applications' is exam-critical. Foundation models (like GPT or Llama) are large pre-trained AI systems that can be adapted for downstream tasks. Sovereign AI refers to a country's capacity to develop, own, and control AI systems — including data, compute, and algorithms — rather than depending on foreign technology stacks.
India's challenge is that global foundation models are trained predominantly on English-language data and Western contexts, making them poorly suited for India's 22 scheduled languages, agrarian economy, and informal healthcare systems. The strategic pivot is toward 'fine-tuning' these base models on Indian datasets — or building indigenous models — and then deploying them through government portals, Jan Dhan accounts, or ASHA worker interfaces.
The Ministry of Electronics and Information Technology (MeitY) oversees AI policy through the IndiaAI Mission. The Digital India Corporation operates as the implementing agency. India's AI governance framework is still evolving — unlike the EU AI Act, India has not yet passed binding AI legislation, relying instead on advisory frameworks from the IT Ministry.
For UPSC, this connects to GS-3 topics: technology policy, digital governance, and India's science and technology strategy.
The distinction between 'foundational models' and 'applications' is exam-critical. Foundation models (like GPT or Llama) are large pre-trained AI systems that can be adapted for downstream tasks. Sovereign AI refers to a country's capacity to develop, own, and control AI systems — including data, compute, and algorithms — rather than depending on foreign technology stacks.
India's challenge is that global foundation models are trained predominantly on English-language data and Western contexts, making them poorly suited for India's 22 scheduled languages, agrarian economy, and informal healthcare systems. The strategic pivot is toward 'fine-tuning' these base models on Indian datasets — or building indigenous models — and then deploying them through government portals, Jan Dhan accounts, or ASHA worker interfaces.
The Ministry of Electronics and Information Technology (MeitY) oversees AI policy through the IndiaAI Mission. The Digital India Corporation operates as the implementing agency. India's AI governance framework is still evolving — unlike the EU AI Act, India has not yet passed binding AI legislation, relying instead on advisory frameworks from the IT Ministry.
For UPSC, this connects to GS-3 topics: technology policy, digital governance, and India's science and technology strategy.
Remember + Why it matters
The key recall facts and exact examiner angle for UPSC CSE are in the Crux app.
01
Key figure and date from this topic
02
Specific number or threshold to remember
03
Policy or regulatory implication
Read + Understand free forever · 30-day free trial