Loading…
Artificial Intelligence is rapidly transforming India's national security architecture, diplomatic functioning, and public administration, creating both unprecedented opportunities and serious governance dilemmas.
The IndiaAI Mission (March 2024) with a ₹10,371 crore outlay marks India's strategic bet on becoming an AI-first nation, with seven pillars including compute, datasets, and safe & trusted AI.
Algorithmic decision-making in security and welfare raises constitutional concerns under Articles 14, 19, and 21, particularly after the Puttaswamy (2017) judgment establishing privacy as a fundamental right.
India faces a 'regulatory trilemma' — balancing innovation, sovereignty, and rights — while choosing between the EU's rights-based, US's market-led, and China's state-controlled AI models.
Without embedding empathy, contextual nuance, and human oversight, AI-driven bureaucracies risk hollowing out the deliberative core of democratic governance.
India's leadership of the Global Partnership on AI (GPAI) and its 'AI for All' vision offer an opportunity to shape inclusive, Global South-friendly AI governance norms.
AI-driven state action implicates the Fundamental Rights triad — Article 14 (equality, non-arbitrariness), Article 19 (free speech and movement), and Article 21 (life, liberty, privacy). The Puttaswamy (2017) judgment requires any privacy intrusion to satisfy the four-part test: legality, legitimate state aim, proportionality, and procedural safeguards. Algorithmic decisions in welfare exclusion, predictive policing, or visa adjudication often fail the proportionality test — they are opaque, non-reviewable, and lack due-process guarantees. The Anuradha Bhasin (2020) ruling on Section 144 and the Justice K.S. Puttaswamy (Aadhaar) (2018) judgment on biometric proportionality offer the doctrinal scaffolding India urgently needs to apply to AI.
AI is rewriting the grammar of national security. Autonomous drone swarms, AI-enabled cyber operations, deepfake-driven information warfare, and algorithmic SIGINT analysis are now central to capability planning. India's adversaries — China's 'intelligentised warfare' doctrine and Pakistan's expanding cyber units — make AI catch-up an existential imperative. Yet, as Raja Mohan warns, ceding security judgment to algorithms risks strategic blunders: AI cannot read the silence of a Wagah-border handshake, the tactical pause of a diplomat, or the cultural nuance of a Galwan-type standoff. The CDS, theatre commands, and DRDO's Centre for Artificial Intelligence and Robotics (CAIR) must embed AI as a decision-support, not a decision-replacement, system.
AI governance straddles the Union List (Defence, Foreign Affairs, Communications), State List (Public Order, Police), and Concurrent List (Criminal Law). States like Telangana (T-AIM), Karnataka, and Tamil Nadu have launched independent AI policies, while predictive policing operates without uniform legal standards. This creates a federal patchwork resembling the GST era pre-2017. The absence of a national AI Authority — analogous to the Data Protection Board under DPDP Act 2023 — leads to regulatory arbitrage and rights asymmetry. Cooperative federalism, invoked in the Sarkaria (1988) and Punchhi (2010) Commission reports, must extend to algorithmic governance through an inter-state council mechanism.
AI has become the new currency of geopolitical influence. India's chairmanship of the Global Partnership on AI (GPAI) in 2024, the New Delhi Declaration, and its leadership at the Bletchley (2023) and Seoul (2024) AI Safety Summits position it as a bridge between the Global North's regulatory rigour and the Global South's developmental priorities. The US-India iCET (initiative on Critical and Emerging Technologies) and the EU-India Trade and Technology Council (TTC) reflect strategic alignment, but India must avoid 'digital colonialism' — where foundational models, training data, and standards remain western-controlled. Sovereign AI — Indic-language LLMs like BharatGPT, AI4Bharat, and Sarvam — is therefore a strategic, not merely cultural, imperative.
The deepest concern is moral, not technical. Algorithmic bureaucracy, by design, optimises for efficiency, scale, and pattern-recognition — values orthogonal to empathy, discretion, and forgiveness, which are constitutive of democratic statecraft. Welfare exclusion of ineligible-flagged Aadhaar beneficiaries, wrongful arrests via facial misrecognition (documented in the US Robert Williams case, 2020), and the dehumanisation of consular and refugee adjudication illustrate the danger. Gandhian ethics of 'antyodaya' (uplifting the last person) and Ambedkar's constitutional morality demand that AI systems incorporate explainability (XAI), human-in-the-loop oversight, and right to algorithmic appeal — not as add-ons, but as design principles.
AI is projected to add $957 billion to India's economy by 2035 (Accenture-NASSCOM). It can transform agriculture (precision farming reducing input costs by 15-20%), healthcare (AIIMS-tele-radiology), and education (NEP 2020's adaptive learning vision). Yet, automation threatens 30-40% of routine jobs (NITI Aayog estimates), disproportionately affecting India's informal sector and women workers. The IndiaAI Mission's FutureSkills pillar and Skill India 2.0 must be scaled tenfold to manage this transition. Without redistributive policy — universal basic services, reskilling guarantees, and progressive taxation of AI-derived productivity gains — AI risks deepening inequality, contradicting the Directive Principles under Articles 38 and 39.
India has one of the world's lowest civil-servant-to-citizen ratios (~16 per 1,000 vs OECD avg of ~70). AI can multiply state capacity in tax administration (GSTN's AI-driven fraud detection recovered ₹2.1 trillion in evasion 2017-23), customs (AI-led risk profiling at JNPT cut clearance time by 40%), and judicial backlog management (SUPACE and SUVAS systems in the Supreme Court). Without AI, governance simply cannot scale.
China's 'New Generation AI Development Plan' (2017) targets global AI leadership by 2030; PLA's 'intelligentisation' doctrine is operational. The Russia-Ukraine war (2022-) and Israel-Hamas conflict (2023-) revealed AI-driven targeting (Lavender, Gospel systems). Falling behind is not a policy choice — it is an invitation to strategic vulnerability. India's CAIR, AI-enabled NETRA, and AIDeX programmes are minimum deterrent investments.
AI can operationalise 'AIforAll'. Krishi-DSS provides AI-powered advisories to 100 million farmers; Bhashini's AI translation supports 22 scheduled languages, narrowing the digital divide; AI-driven TB detection (Wadhwani AI) has screened 10+ million Indians. For a country of 1.4 billion, where governance gaps cost lives, refusing AI on theoretical risk grounds is itself unethical.
Algorithmic decisions in welfare (Telangana's Samagra Vedika), policing (Delhi Police's AFRS arrests at protests), and revenue (income-tax faceless assessment errors) often lack reasoned orders, audit trails, and effective appeal — violating Article 14's non-arbitrariness mandate (Maneka Gandhi, 1978) and natural justice. The Aadhaar judgment (2018) explicitly demanded proportionality; current AI deployments fail this test.
Real-time facial recognition, CCTNS integration, social-media monitoring (Project Insight), and AI-driven OSINT can produce a panopticon effect. Studies (Pew, MIT Media Lab) show citizens self-censor under surveillance — directly undermining Article 19(1)(a). India lacks a surveillance reform law analogous to the US Foreign Intelligence Surveillance Act (FISA) or the UK Investigatory Powers Act, leaving executive discretion unchecked.
Generative AI hallucinates (ChatGPT fabricated case citations in Mata v. Avianca, 2023). Facial recognition has 10-100x higher error rates for women and darker-skinned individuals (NIST 2019, Gender Shades study). When an algorithm denies a ration card or flags an innocent traveller, who is liable — the vendor, the officer, the ministry? India has no clear algorithmic accountability statute, leaving citizens remediless.
Equality before law (non-arbitrary algorithms), freedoms of speech-movement-trade, and right to life and privacy. Puttaswamy (2017) made privacy a fundamental right; proportionality is mandatory for AI-driven state action.
India's first horizontal data protection law. Establishes Data Protection Board, consent framework, data principal rights, and significant data fiduciary obligations — foundational for AI training data governance.
Section 69 (interception), Section 79 (intermediary safe harbour), and the 2023 amendment requiring AI/deepfake content labelling and fact-checking unit oversight (currently sub-judice in Bombay HC).
Modernises interception powers, governs digital communication infrastructure on which AI services run, and replaces colonial-era Telegraph Act 1885.
Covers organised cybercrime, hate speech, and acts endangering sovereignty — directly applicable to AI-generated deepfakes and disinformation.
Will replace IT Act 2000; expected to include dedicated chapters on AI regulation, high-risk AI systems, and platform accountability.
Nodal ministry for AI policy, IndiaAI Mission implementation, and DPDP Act administration
Executes the 7-pillar IndiaAI Mission — compute, datasets, innovation centre, application development, FutureSkills, startup financing, safe & trusted AI
Authored National Strategy for AI (2018) and AI for All; coordinates use-case adoption across ministries and states
Statutory body under DPDP Act 2023 to adjudicate data breach complaints and impose penalties up to ₹250 crore
Recommended a domestic AI law and a statutory AI authority (July 2023 recommendations on regulating AI)
Develops AI/ML solutions for the armed forces — autonomous platforms, secure communications, command & control systems
Multilateral forum (29 member countries) on responsible AI; India was Lead Chair 2024 and hosts AI Safety Institute conversations
| Parameter | European Union | United States | China | India (Emerging) |
|---|---|---|---|---|
| Regulatory Philosophy | Rights-based, precautionary | Market-led, sector-specific | State-controlled, party-led | Innovation-first, light-touch (evolving) |
| Flagship Law/Policy | EU AI Act (entered force August 2024) | Executive Order 14110 (Oct 2023); Biden-Trump policy shifts 2025 | Algorithmic Recommendation Regs (2022); Generative AI Measures (2023) | DPDP Act 2023; IndiaAI Mission 2024; Digital India Act (proposed) |
| Approach to High-Risk AI | Risk-tiered (unacceptable, high, limited, minimal) | Voluntary safety commitments + NIST AI RMF | Algorithmic registry, security review, content controls | No statutory classification yet; advisory framework under MeitY |
| Biometric Surveillance | Real-time public biometric ID largely banned | City/state-level patchwork (some bans) | Pervasive state deployment (Sharp Eyes, Skynet) | Widely deployed (NAFRS, state AFRS); no dedicated statute |
| Strategic Goal | Trustworthy AI; Brussels Effect on global norms | Maintain technological supremacy over China | Global AI leader by 2030; 'AI with Chinese characteristics' | AI for All; Global South leadership; sovereign AI stack |
Telangana's Samagra Vedika integrates 25+ government databases using AI/ML to detect ineligible beneficiaries across schemes like Aasara pensions, ration cards, and Rythu Bandhu.
By 2022, the system flagged and removed lakhs of 'duplicate' or 'ineligible' beneficiaries. Civil society audits (Internet Freedom Foundation, 2022) found significant false-positive rates, with genuinely poor citizens losing entitlements without notice or hearing.
While the state claimed savings of ₹500+ crore, several beneficiaries had to approach courts or media to restore benefits. Cases highlighted absence of statutory appeal, explainability, and audit.
Algorithmic welfare governance must embed Article 14's procedural due process — reasoned orders, prior notice, and accessible grievance redressal — failing which efficiency gains come at constitutional cost.
When COVID-19 cancelled UK A-Level exams, Ofqual deployed an algorithm to estimate grades, factoring in school history. Nearly 40% of grades were downgraded from teacher predictions.
The algorithm systematically disadvantaged students from state schools and disadvantaged backgrounds, advantaging private schools. Public protests (#FuckTheAlgorithm) and judicial review threats followed.
The UK government withdrew the algorithm within 4 days, reverting to teacher-predicted grades. The Education Secretary apologised; the Ofqual chief resigned.
Even technically 'accurate' algorithms can encode structural inequality. High-stakes algorithmic decisions need impact assessments, public consultation, and rapid override mechanisms — principles India must internalise as it scales AI in education, exams (e.g., NTA), and welfare.
Bhashini, the National Language Translation Mission, builds open-source AI models for translation, speech-to-text, and TTS across 22 scheduled languages, addressing India's deepest digital divide — language.
Integrated with UMANG, MyGov, e-Sansad, and Digital India Bhashini app. Open-sourced datasets via AI4Bharat (IIT Madras) and NPCI's UPI-Bhashini voice payments.
Over 30+ million translations rendered (2024); voice-based UPI piloted in Hindi/Tamil/Telugu. Bhashini represents a public-good AI model — locally developed, multilingual, open-source.
Sovereign, open, and inclusive AI is feasible. India's DPI-led approach (Aadhaar → UPI → ONDC → Bhashini) shows AI can be designed for equity, not just efficiency — a model the Global South can adopt.
AI is not a tool India can choose to adopt or reject — it is the operating environment in which the 21st-century state will function. The real question, as Raja Mohan's editorial poignantly observes, is whether India will let algorithms hollow out the empathy, deliberation, and constitutional morality that define its democracy, or whether it will design AI systems that amplify these very values. The IndiaAI Mission, DPDP Act, GPAI leadership, and DPI legacy offer a foundation — but only a foundation. What India now needs is the institutional architecture, legal rigour, and moral imagination to build, on that foundation, the world's first AI-augmented constitutional republic. The choice between sovereign-democratic AI and dependent-technocratic AI will shape not just India's destiny, but the global future of governance itself.
AI is the natural fourth layer over India's DPI stack — Identity (Aadhaar) → Payments (UPI) → Commerce (ONDC) → Intelligence (AI). The Puttaswamy (Aadhaar) judgment's proportionality test must extend to AI deployments.
AI-enabled cyber threats (deepfakes, automated phishing, AI malware) require parallel investment in AI defence — CERT-In, NCIIPC, and the National Cyber Security Strategy must integrate AI threat intelligence.
AI is central to QUAD's Critical & Emerging Technologies Working Group, US-India iCET, and counter-China tech containment. India's positioning here shapes its strategic autonomy.
AI training has massive energy/water footprint (GPT-3 training ~1,287 MWh, 552 tonnes CO2). India's Net Zero 2070 goal requires 'Green AI' standards in data centres and compute.
AI policy spans Union, State, and Concurrent Lists. The Sarkaria (1988) and Punchhi (2010) Commission frameworks for cooperative federalism need a 21st-century algorithmic update via Inter-State Council.
“Critically examine the implications of Artificial Intelligence on India's national security architecture and diplomatic functioning. Suggest a framework for sovereign, human-centric AI governance. (250 words)”
“"Without empathy and contextual nuance, AI-driven bureaucracy is a dangerous turn for democratic governance." Discuss in the context of India's emerging AI ecosystem. (250 words)”
“Compare and contrast the AI governance approaches of the European Union, the United States, and China. What lessons can India draw to craft its own model? (250 words)”
GS2 > Governance, E-Governance; GS3 > Science & Tech, Internal Security; GS4 > Ethics, Technology and Society
Current Events, Science & Technology, Polity (Fundamental Rights)
DAF-based + emerging tech personality test theme
General Awareness — Schemes, Polity
General Awareness — Digital governance, RBI's AI/ML use
Introduce the concept of Artificial Intelligence (AI). How does AI help clinical diagnosis? Do you perceive any threat to privacy of the individual in the use of AI in healthcare?
What are the main socio-economic implications arising out of the development of IT industries in major cities of India?
Implementation of Information and Communication Technology (ICT) based projects/programmes usually suffers in terms of certain vital factors. Identify these factors and suggest measures for their effective implementation.
With reference to Web 3.0, consider the following statements... (related to decentralised tech, AI, semantic web).
Local self-government can be best explained as an exercise in
Answer: Democratic decentralisation
Mains: 2023 (AI in healthcare), 2020 (IT industry), 2019 (ICT projects), 2018 (technology policy). Prelims: 2022, 2020, 2018 — appearing almost yearly since 2018 in some form.
₹10,371 crore comprehensive national programme (March 2024) with 7 pillars to build India's AI ecosystem — compute infrastructure, datasets, innovation centres, applications, skills, startup funding, and safe & trusted AI.
Four-part test for any state action infringing privacy: (1) legality, (2) legitimate state aim, (3) proportionality (rational nexus, necessity, balance), (4) procedural safeguards.
Emerging legal doctrine extending constitutional principles — equality, due process, accountability — to algorithmic and AI-driven state decision-making.
Global Partnership on Artificial Intelligence — 29-member multilateral initiative hosted at OECD Paris; promotes responsible AI development; India was Lead Chair 2024.
AI capabilities (foundational models, compute, data) developed and controlled domestically to ensure strategic autonomy, cultural-linguistic relevance, and freedom from foreign tech dependence.
Design principle requiring meaningful human oversight, review, and override of algorithmic decisions — especially in high-stakes domains like security, welfare, and justice.
November 2023 declaration signed by 28+ countries (including India, US, China, EU) at UK's AI Safety Summit committing to cooperation on frontier AI safety.