Why this matters now
AI will be the defining technology of the next decade. Whether India joins US and China as AI superpowers or remains a second-tier player depends on policy choices in the next 3-5 years. The IndiaAI Mission is the most ambitious AI policy intervention India has made.
IndiaAI Mission 2024
Approved March 2024, ₹10,372 crore over five years. Seven pillars:
- IndiaAI Compute Capacity — PPP for 10,000+ GPU infrastructure;
- IndiaAI Innovation Centre — foundational LLMs in Indian languages, multimodal AI, sovereign capabilities;
- IndiaAI Datasets Platform — non-personal datasets for AI development;
- IndiaAI Application Development Initiative — healthcare, education, agriculture, governance applications;
- IndiaAI FutureSkills — UG/PG/PhD courses; Data and AI labs in Tier 2/3 cities;
- IndiaAI Startup Financing — deep-tech focus;
- Safe & Trusted AI — guidelines, frameworks, AI Safety Institute.
Implementing agency: IndiaAI Independent Business Division within Digital India Corporation under MeitY.
NITI Aayog National AI Strategy 2018
NITI Aayog's 'National Strategy for Artificial Intelligence' (June 2018) — 'AI for All' — was India's first comprehensive AI policy framework. Five focus sectors:
- Healthcare;
- Agriculture;
- Education;
- Smart Cities & Infrastructure;
- Smart Mobility & Transportation.
Subsequent NITI papers on Responsible AI (2021), AI for Education (2022), GenAI (2023). NITI's role gradually transferred to MeitY/IndiaAI Mission.
EU AI Act vs India's approach
| Dimension | EU AI Act | India |
|---|---|---|
| Architecture | Comprehensive horizontal regulation | Light-touch, sector-specific |
| Risk categorisation | Four risk tiers (banned, high, limited, minimal) | No formal tiering |
| Penalties | Up to 7% global turnover or €35 mn | Sector-specific, lower |
| Banned uses | Social scoring, real-time biometric ID (with exceptions), predictive policing | No comprehensive ban list |
| High-risk applications | Heavy compliance obligations | No high-risk category |
| Transparency | Deepfakes labelled; chatbots disclosed | IT Rules 36-hour takedown for deepfakes |
| Foundation models | Documentation, transparency obligations | No specific obligations |
| Status | In force August 2024, applies August 2026 | No comprehensive AI Act yet |
India's approach: innovation over restriction; concerned that heavy regulation could constrain Indian AI startups; sector-specific better than horizontal; international cooperation through GPAI.
DPDP Act 2023 intersection
The Digital Personal Data Protection Act 2023 has significant intersection with AI:
- Defines personal data — AI training data must comply;
- Consent-based processing — affects AI personalisation;
- Synthetic data tagged as personal data under certain interpretations;
- Significant Data Fiduciaries (large AI platforms) face additional obligations;
- Data Principal rights (access, correction, erasure) affect AI memory and model retraining;
- Cross-border transfer framework — affects training data flow.
The DPDP Rules (under preparation) will clarify many AI-relevant provisions.
Deepfake regulation
India saw major incidents: Rashmika Mandanna deepfake (November 2023); Katrina Kaif, Kajol deepfakes; multiple political deepfakes during 2024 elections.
India's response:
- IT Rules 2021 — platforms must take down deepfake content within 36 hours of notification;
- MeitY advisories March 2024 (initially mandated approval for under-tested AI; rolled back after backlash);
- IT Act sections 66E (privacy), 66D (impersonation cheating), 67/67A (obscene content);
- BNS 2023 sections on defamation, criminal intimidation;
- Election Commission's Voluntary Code of Ethics + advisories;
- Comprehensive Digital India Act under preparation.
Proposed measures: mandatory provenance/watermarking of AI content; industry self-regulation; education and media literacy.
GPAI and global AI forums
Global Partnership on AI (GPAI) — launched June 2020 by G7 + India + Korea + Mexico + Slovenia + EU. India:
- Founding member;
- Hosted GPAI Annual Summit New Delhi December 2023;
- Chaired GPAI 2024 (succeeding Japan);
- Champions inclusive AI for developing countries.
Other forums:
- G7 Hiroshima AI Process (2023);
- UN AI Advisory Body;
- AI Safety Summit Bletchley Park 2023, Paris February 2025;
- OECD AI Principles.
India's strategy: use multilateral forums to shape global AI norms in favour of inclusive, accessible, multi-polar AI development.
AI startup ecosystem
India's AI startup landscape (2024):
- Bharat GPT (Reliance Jio + IIT) — LLMs for Indian languages;
- Krutrim (Ola founder Bhavish Aggarwal) — multilingual LLMs; unicorn status;
- Sarvam AI — open-source Indian language LLMs;
- Yotta Data Services — AI infrastructure;
- Avant Garde Innovations, Niki AI, Lumina Datamatics, others;
- ~500+ AI startups (NASSCOM 2024 estimate);
- ~$8 billion cumulative funding 2014-24.
Global tech in India: Microsoft, Google, Amazon, Meta, Anthropic, OpenAI all major presences; Microsoft India + OpenAI partnership.
India's strategic AI advantages
- Talent — 5 lakh+ AI researchers; world's largest STEM graduate pool; English depth;
- Identity infrastructure — Aadhaar 140 cr enrolments;
- Multilingual — 22 official + 1,600+ dialects; non-English AI critical;
- Consumer demand — 800+ mn internet users; massive Hindi/Indian language content;
- DPI architecture — UPI, ONDC, AA enable AI deployment;
- Government support — IndiaAI Mission, PLI, Make in India;
- Foreign capital — Microsoft/Google/Amazon/Meta investments;
- Diaspora leadership — Sundar Pichai (Google), Satya Nadella (Microsoft), Aravind Srinivas (Perplexity).
India's strategic AI challenges
- GPU access — US export controls restrict advanced AI chips;
- Foundation model gap — Indian models early-stage vs OpenAI/Google/Anthropic/Meta;
- Training data quality — multilingual data sparse, biased;
- Research funding — government R&D ~0.65% of GDP vs USA 3.5%, China 2.5%;
- Talent retention — best minds drift to US;
- Regulatory ambiguity — comprehensive AI law absent;
- Skilling — workforce transition;
- Energy — AI training needs massive electricity.
Strategic positioning: application leader rather than foundation model leader; Indian language AI as differentiator; DPI-AI integration; public+private partnership; Global South solidarity for inclusive AI.
"India's AI moment will be defined by whether it can leverage its DPI advantage to deploy AI at scale, rather than competing head-on with US foundation model leadership." — paraphrasing Nandan Nilekani's recurring framing
UPSC PYQs and likely future questions
UPSC angle
AI policy questions span GS-3 (Science & Tech, Economy, Governance). Strong answers describe IndiaAI Mission accurately, address regulatory approach comparison, and connect to DPI architecture.
- 2023 GS-3: "Discuss the IndiaAI strategy. What are the components and challenges?"
- 2024 GS-3: "Compare India's AI regulatory approach with the EU AI Act. Which model is more suitable for Indian conditions?"
- Likely 2026: "Examine the deepfake challenge for Indian democracy. What policy frameworks can address it?"
- Likely 2026: "Discuss India's role in global AI governance through GPAI. To what extent does India advance Global South interests?"
S&T cluster — 3/4
Paired with India Stack and Semiconductor Mission. One more piece (Space Programme) will close the cluster.