Keeping humanity at the centre of the AI revolution

Keeping Humanity at the Centre of the AI Revolution

UPSC Prelims + Mains Study Notes | GS-III / GS-IV | Science & Technology / Ethics

1. At a Glance

• AI governance is the body of principles, institutional mechanisms, and legal frameworks designed to ensure that Artificial Intelligence development serves human dignity, equity, and accountability rather than autonomous algorithmic logic. [1][2]
• The core tension: AI's transformative potential in healthcare, education, economic inclusion, and disaster management vs. risks of job displacement, algorithmic bias, surveillance, and erosion of human agency. [4]
• Why UPSC-critical: Bridges GS-III (Science & Technology, Digital Economy) and GS-IV (Ethics, Integrity, Accountability); directly tested in essay and Mains opinion-type questions since 2022.
• India is among the earliest large democracies to release national AI Governance Guidelines (Nov 2025) and launch a PM-level vision framework (M.A.N.A.V.), making this a live policy topic. [3][5]

2. Why in the News

• India AI Impact Summit 2026 (Feb 19, 2026): PM Narendra Modi outlined the M.A.N.A.V. (Moral & Ethical Systems, Accountable Governance, National Sovereignty, Inclusive Development, Viability through Trust) framework for human-centric AI. [5]
• MeitY released India AI Governance Guidelines (November 5, 2025) — a comprehensive national framework for safe, responsible, inclusive AI adoption. [3]
• 4th UNESCO Global Forum on the Ethics of AI (GFEAI 2026) announced for September 14–17, 2026, Riyadh, Saudi Arabia; theme: "Transforming Global Cooperation for Ethical AI Governance." [1]
• UN Global Dialogue on AI Governance scheduled for July 6–7, 2026, Geneva — co-organised by UNESCO and ITU to ensure AI governance reflects all nations, not just technologically advanced ones. [1]
• Op-ed by Ashwani Kumar (Senior Advocate, Supreme Court; former Union Law Minister) in The Hindu, June 26, 2026, called for keeping human dignity and ethics at the centre of the AI revolution. [4]

3. Background & Evolution

4. Core Static Facts

India-specific:

• IndiaAI Mission: Cabinet-approved outlay of ₹10,372 crore (~₹10,300+ crore); covers compute, data, skilling, innovation, and governance pillars. [3]
• Implementing Ministry: Ministry of Electronics and Information Technology (MeitY); mission anchored under IndiaAI. [3]
• India AI Governance Guidelines (Nov 2025) — structured around:
• 7 Guiding Sutras (principles for ethical/responsible AI)
• 6 Pillars of AI Governance (recommendations)

• Short, medium, long-term Action Plan

• New institutions established: AI Governance Group; Technology & Policy Expert Committee; AI Safety Institute (India). [3]

• High-risk AI systems: not permitted unrestricted deployment under the Guidelines; risk-based, evidence-led, proportional approach mandated. [3]
• IT Amendment Rules 2026: formal definition and regulation of synthetically generated content (deepfakes). [3]

M.A.N.A.V. Framework (PM Modi, Feb 19, 2026): | Letter | Pillar | |--------|--------| | M | Moral and Ethical Systems | | A | Accountable Governance | | N | National Sovereignty | | A | Inclusive Development (Antyodaya) | | V | Viability through Trust |

Global:

• UNESCO Recommendation on Ethics of AI (2021): adopted by 193 member states; first-ever global standard. [1]
• Readiness Assessment Methodology (RAM): UNESCO tool; over 75 countries across 5 continents have initiated/completed RAM assessments as of October 2025. [1]
• Global Civil Society & Academic Network on AI Ethics: launched June 2025 at 3rd Global Forum on Ethics of AI, Thailand. [1]
• MeitY–UNESCO Multi-Stakeholder Consultation on Safety and Ethics in AI: hosted jointly (PIB, 2024). [6]

5. Multi-Dimensional Analysis

Economic

• AI automates repetitive tasks, threatening routine employment (especially low-skill jobs); simultaneously creates new categories of work requiring AI-augmented skills. [4]
• IndiaAI Mission (₹10,372 cr) targets AI-driven economic growth across agriculture, health, finance, and manufacturing. [3]
• Risk: concentration of AI benefits among Big Tech, widening inequality between AI-capable and AI-excluded nations.

Social

• AI's potential: cancer screening, terminal illness prediction, robotic nursing, targeted aid to marginalised communities, expanded educational access. [4]
• Risk: algorithmic bias can encode and amplify existing caste, gender, and income inequalities in credit scoring, hiring, and law enforcement.
• India's Antyodaya principle in M.A.N.A.V. explicitly links AI to last-mile inclusion. [5]

Geopolitical / Strategic

• AI governance is a new frontier of great-power competition (US, China, EU each developing competing regulatory architectures).
• UN AI Governance Dialogue (Jul 2026) seeks to prevent Global South marginalisation in AI norm-setting. [1]
• National Sovereignty is an explicit pillar of India's M.A.N.A.V. framework, reflecting concerns over data colonialism and foreign AI dependency. [5]

Legal / Constitutional

• IT Amendment Rules 2026 regulate deepfakes — legally defining synthetic media for the first time in India. [3]
• Article 19(1)(a) (free speech) and Article 21 (life and personal liberty/privacy) are implicated by AI surveillance, profiling, and content moderation.
• India lacks a standalone AI Act (unlike the EU AI Act, 2024); governance is presently guideline-based (non-binding). [3]

Scientific / Technological

• AI applications flagged as transformative: disaster management, weather forecasting, environmental sustainability, cancer screening, robotic surgery. [4]
• UNESCO's AI Literacy Training for civil servants is building state capacity for ethical AI deployment globally. [1]
• AI Safety Institutes (India, UK, US) focus on frontier AI risk evaluation — a new institutional form in science governance. [3]

Ethical / Governance

• Core ethical concepts: human dignity, transparency, accountability, non-maleficence, fairness, inclusivity, proportionality. [1][4]
• The UNESCO Recommendation introduces the concept of "AI ethical impact assessment" prior to deployment. [1]
• India's "whole-of-government" model integrates governance into the AI ecosystem from inception, not as an afterthought. [3]
• Calls for a "pause" in AI development by some technology leaders (e.g., 2023 open letter by Elon Musk, Yoshua Bengio et al.) reflect the depth of ethical concern. [4]

6. Recent Developments (Last 12–18 Months)

• November 5, 2025: MeitY releases India AI Governance Guidelines — risk-based, principle-led national framework. [3]
• October 2025: Over 75 countries complete or initiate UNESCO RAM (Readiness Assessment Methodology) for AI ethics implementation. [1]
• June 2025: Global Civil Society & Academic Network on AI Ethics launched at 3rd UNESCO GFEAI, Chiang Mai, Thailand. [1]
• February 19, 2026: PM Modi delivers M.A.N.A.V. framework address at India AI Impact Summit; global leaders emphasise human-centred AI governance for Digital Public Infrastructure. [5]
• 2026: IT (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules 2026 formally regulate synthetically generated content in India. [3]
• September 14–17, 2026 (upcoming): 4th UNESCO GFEAI, Riyadh — "Transforming Global Cooperation for Ethical AI Governance." [1]
• July 6–7, 2026 (upcoming): UN Global Dialogue on AI Governance, Geneva, co-organised by UNESCO & ITU. [1]

7. Prelims Hooks

1. 193 countries adopted the UNESCO Recommendation on the Ethics of AI in 2021 — first global AI ethics standard. [1]
2. The 4th UNESCO Global Forum on the Ethics of AI (2026) will be held in Riyadh, Saudi Arabia, co-hosted with the Kingdom of Saudi Arabia. [1]
3. UNESCO's Readiness Assessment Methodology (RAM) has been initiated or completed by over 75 countries from 5 continents as of October 2025. [1]
4. IndiaAI Mission was approved with an outlay of over ₹10,372 crore (commonly cited as ₹10,300+ crore). [3]
5. India AI Governance Guidelines (Nov 2025) are structured around 7 Guiding Sutras and 6 Pillars — released by MeitY (not NITI Aayog). [3]
6. Three new institutions under India's AI governance architecture: AI Governance Group, Technology & Policy Expert Committee, and AI Safety Institute. [3]
7. M.A.N.A.V. stands for: Moral & Ethical Systems, Accountable Governance, National Sovereignty, Antyodaya (Inclusive Development), Viability through Trust — announced February 19, 2026. [5]
8. India's AI Governance Guidelines adopt a risk-based, evidence-led, proportional approach — high-risk AI systems are not allowed unrestricted deployment. [3]
9. The IT Amendment Rules 2026 (not the IT Act itself) formally define and regulate synthetically generated content (deepfakes) in India. [3]
10. The UN Global Dialogue on AI Governance (Jul 2026, Geneva) is co-organised by UNESCO and the ITU — not the UN General Assembly alone. [1]
11. The OECD Principles on AI (2019) were the first intergovernmental standard on trustworthy AI — predating the UNESCO Recommendation by two years. [2]
12. MeitY and UNESCO jointly hosted a Multi-Stakeholder Consultation on Safety and Ethics in AI in India (2024). [6]

8. Mains Relevance

Plausible Mains Question Stems:

1. "Artificial Intelligence promises transformative benefits but poses existential risks to human dignity and agency. Critically examine India's regulatory response and suggest a framework that keeps humanity at the centre of the AI revolution." (GS-III / GS-IV, 15 marks)
2. "The UNESCO Recommendation on the Ethics of AI (2021) represents an important milestone in global governance, yet remains non-binding. Analyse its significance and limitations in the context of India's AI governance strategy." (GS-II / GS-III, 15 marks)
3. "'AI should be an extension of human aspiration, not a replacement for human judgment.' Evaluate PM Modi's M.A.N.A.V. vision in the light of India's developmental priorities and constitutional values." (GS-IV / Essay)

9. Related Topics to Study Next

10. Common Errors / Trap Areas

1. MeitY vs. NITI Aayog: India AI Governance Guidelines and IndiaAI Mission are under MeitY, not NITI Aayog (though NITI Aayog published earlier strategy papers). Confusing the two is a common mistake.
2. UNESCO Recommendation (2021) is non-binding: It is a recommendation, not a treaty or convention. Do not treat it as legally enforceable international law.
3. India has no standalone AI Act: India's current framework (Nov 2025 Guidelines) is principle-based and non-binding — unlike the EU AI Act which is a regulation. Aspirants often conflate guidelines with legislation.
4. M.A.N.A.V. date: Announced February 19, 2026 at the India AI Impact Summit — not at any UN forum or G20 meeting.
5. RAM = Readiness Assessment Methodology (UNESCO): Sometimes confused with RAM (Random Access Memory in tech context) or other government schemes. In AI governance, RAM is UNESCO's diagnostic tool for member states.