What has govt. laid down on AI labelling?

UPSC Study Note: Government's AI Labelling Rules (IT Amendment Rules, 2026)

1. At a Glance

• MeitY notified the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026 on February 10, 2026; effective February 20, 2026. [S1]
• Rules mandate that social media platforms prominently label AI-generated/synthetically generated content and embed permanent metadata for traceability. [S1][S2]
• Takedown timelines for all unlawful content have been tightened from 24–36 hours to 2–3 hours; non-consensual intimate imagery must be removed within 2 hours. [S3]
• Critical from UPSC perspective: intersection of GS-III (technology governance), GS-II (government policy, intermediary liability), and GS-IV (ethics in AI/media). [S2]

2. Why in the News

• February 10–15, 2026: MeitY notified the 2026 Amendment Rules; drew editorial attention from The Hindu (Feb 15, 2026 Sunday edition) and Business Standard. [S1][S3]
• Triggering factors: rising incidents of AI-generated deepfakes, electoral misinformation, and non-consensual synthetic imagery in India's digital ecosystem. [S2]
• Amendment follows an October 2025 draft/explanatory note that was open for public consultation before finalisation. [S4][S5]
• India's AI Governance Guidelines (enabling safe and trusted AI innovation) were also published around the same time (February 2026), signalling a coordinated regulatory push. [S6]

3. Background & Evolution

• Predecessor: IT Rules 2021 (Rule 3 & Rule 4 — due diligence, significant social media intermediary obligations).
• Related: MeitY's 2023 advisories on deepfakes; Telecom Act 2023; Digital Personal Data Protection Act 2023.

4. Core Static Facts

• Full Name: Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026
• Gazette Notification: G.S.R. 120(E), dated February 10, 2026 [S1]
• Effective Date: February 20, 2026 [S1]
• Parent Legislation: Information Technology Act, 2000 (Section 87) [S1]
• Implementing Ministry: Ministry of Electronics and Information Technology (MeitY) [S1]
• Key Regulatory Target: Significant Social Media Intermediaries (SSMIs) — platforms with >5 million users [S2]
• Definition — Synthetically Generated Information (SGI): Any audio, visual, or audio-visual content artificially or algorithmically created, generated, modified, or altered using a computer resource in a manner that it appears real, authentic, or true [S5][S1]
• Exemptions from SGI definition: Smartphone-clicked photographs with minor AI-assisted edits; content edited in "good faith" using AI tools [S2][S1]
• Takedown timelines (post-amendment):
• General unlawful content: 3 hours (was 24–36 hours) [S3]
• Non-consensual intimate imagery (NCII): 2 hours (was 24 hours) [S3]
• Key new obligation — Rule 4(1A): SSMIs must (a) obtain a user declaration for AI-generated content and (b) conduct technical verification before publishing SGI [S5][S2]
• Metadata requirement: Platforms must embed permanent metadata or identifiers to trace the origin of AI-generated content [S1][S2]
• Stated rationale: Counter deepfakes, misinformation, privacy violations, and threats to national integrity [S2][S4]

5. Multi-Dimensional Analysis

Legal / Constitutional

• Amendments derive from Section 87 of the IT Act, 2000 (rule-making power of Central Government). [S1]
• Tightening intermediary obligations raises questions about safe harbour under Section 79 of the IT Act — platforms could lose immunity if they fail to comply. [S3]
• The October 2025 draft's wider SGI definition (encompassing any AI-modified content) was narrowed in the final rules — indicative of industry pushback and legal concerns around over-breadth. [S2][S4]

Scientific / Technological

• Technical verification of user declarations for SGI requires platforms to deploy AI-detection tools — a fast-evolving and imperfect technology. [S5]
• Metadata embedding (e.g., C2PA standards, watermarking) is mandated for traceability; interoperability across platforms remains a challenge. [S1][S2]
• Exemptions for "good faith" AI editing create ambiguity — no technical bright line distinguishes routine filters from synthetic generation. [S2]

Ethical / Governance

• Balances freedom of expression (Article 19(1)(a)) against harms from deepfakes and misinformation. [S2]
• Risk of chilling effect on satire, parody, and political commentary if labelling obligations are broadly interpreted. [S3]
• The 2-hour NCII takedown window reflects a gender-sensitive governance approach given disproportionate targeting of women by synthetic intimate imagery. [S3]

Economic / Administrative

• Significant compliance burden on SSMIs — investment in detection infrastructure, moderation capacity, and audit systems. [S3][S2]
• Smaller platforms approaching the 5-million threshold may face incentive to remain below it to avoid SSMI obligations. [S3]
• MeitY simultaneously issued India AI Governance Guidelines (Feb 2026) signalling a dual track: rules for platforms + enabling framework for AI innovation. [S6]

Geopolitical / Strategic

• Part of a global trend: EU AI Act (2024), China's deep synthesis regulations (2022), and US executive orders on AI (2023–24) all address synthetic content. [S2]
• India's rules are among the strictest globally on takedown timelines, signalling assertive digital sovereignty. [S3]

6. Recent Developments (Last 12–18 Months)

• October 22, 2025: MeitY released draft SGI labelling amendment + Explanatory Note; public consultation initiated. [S4][S5]
• October 2025: MeitY FAQ document on proposed SGI rules published. [S4]
• February 10, 2026: Final rules notified (G.S.R. 120(E)). [S1]
• February 2026: India AI Governance Guidelines ("Enabling Safe and Trusted AI Innovation") released by PIB/MeitY. [S6]
• February 20, 2026: Amendment Rules came into force. [S1]
• April 2026: Business Standard reports ongoing industry debate on compliance costs and definitional ambiguity of SGI; MeitY Secretary clarified amendments are not linked to specific content takedown controversies. [S2][S8]

7. Prelims Hooks

1. The IT Amendment Rules, 2026 were notified by MeitY on February 10, 2026 via Gazette Notification G.S.R. 120(E). [S1]
2. Rules apply to Significant Social Media Intermediaries (SSMIs) — platforms with more than 5 million users in India. [S2]
3. SGI = Synthetically Generated Information: audio/visual/audio-visual content algorithmically created to appear real or authentic. [S5]
4. General unlawful content takedown window: reduced to 3 hours (previously 24–36 hours). [S3]
5. Non-consensual intimate imagery must be taken down within 2 hours under the new rules. [S3]
6. The October 2025 draft had a wider SGI definition than the final February 2026 rules — the final version carved out exemptions for routine AI-assisted editing. [S2][S4]
7. Rule 4(1A) explicitly mandates SSMIs to obtain user declaration and conduct technical verification before publishing AI-generated content. [S5]
8. Enabling Act: Information Technology Act, 2000, Section 87 (rule-making power). [S1]
9. The rules require platforms to embed permanent metadata/identifiers in AI-generated content for origin traceability. [S1]
10. MeitY cited deepfakes, misinformation, privacy violations, and national integrity threats as the four core rationales in its explanatory document. [S4]
11. Rules came into force on February 20, 2026 — 10 days after notification. [S1]
12. The SGI labelling requirement was first proposed in October 2025 before final notification. [S2]
13. India's AI Governance Guidelines (separate from intermediary rules) were also released in February 2026 by PIB. [S6]

8. Mains Relevance

Plausible Mains Questions: 1. "The IT (Intermediary Guidelines) Amendment Rules, 2026 seek to balance freedom of expression with the menace of synthetic media. Critically analyse the regulatory approach and its limitations." (GS-II/GS-IV) 2. "Discuss the governance challenges posed by AI-generated content in India's information ecosystem. How do the 2026 AI labelling rules address these, and what gaps remain?" (GS-III) 3. "Examine how India's approach to regulating deepfakes and synthetically generated information compares with global frameworks such as the EU AI Act." (GS-II/GS-III)

9. Related Topics to Study Next

10. Common Errors / Trap Areas

1. Wrong ministry: Do not confuse MeitY with the Ministry of Information & Broadcasting (MIB) — MIB regulates broadcast/OTT content under IT Rules Part III; MeitY handles intermediary obligations under Part II. [S1]
2. Wrong effective date: Rules were notified on Feb 10 but effective from Feb 20, 2026 — MCQs may test this 10-day gap. [S1]
3. SGI definition scope: The draft (October 2025) definition was broader; the final rules have exemptions for routine AI-assisted editing. Do not use the draft definition for the enacted rule. [S2][S4]
4. Takedown timelines — two different windows: General unlawful content = 3 hours; NCII specifically = 2 hours. Conflating them is a common error. [S3]
5. SSMI threshold confusion: The 5-million user threshold applies to SSMIs; the labelling/verification obligations under Rule 4(1A) apply only to SSMIs, not all intermediaries. [S2][S5]

11. Sources

• [S1] The Information Technology (Intermediary Guidelines) Amendment Rules, 2026 — Gazette — https://www.meity.gov.in/static/uploads/2026/02/550681ab908f8afb135b0ad42816a1c9.pdf — (Tier 1)
• [S2] Business Standard — "AI labels on content: Curbing misinformation or adding compliance load?" — https://www.business-standard.com/technology/tech-news/ai-contents-labels-india-meity-misinformation-compliance-digital-media-126042400645_1.html — (Tier 4)
• [S3] Business Standard — "Govt toughens rules on AI content, mandates 3-hour takedown timeline" — https://www.business-standard.com/technology/tech-news/it-ministry-shortens-takedown-timelines-intermediaries-ai-content-rules-126021001625_1.html — (Tier 4)
• [S4] MeitY Explanatory Note (October 22, 2025) — https://www.meity.gov.in/static/uploads/2025/10/8e40cdd134cd92dd783a37556428c370.pdf — (Tier 1)
• [S5] MeitY FAQ on Proposed SGI Rules — https://www.meity.gov.in/static/uploads/2025/10/065b6deb585441b5ccdf8be42502a49c.pdf — (Tier 1)
• [S6] India AI Governance Guidelines (PIB, February 2026) — https://static.pib.gov.in/WriteReadData/specificdocs/documents/2026/feb/doc2026215790801.pdf — (Tier 1)
• [S7] MeitY Advisory to Intermediaries (PIB) — https://www.pib.gov.in/PressReleaseIframePage.aspx?PRID=1990542 — (Tier 1)
• [S8] Business Standard — "Draft IT rules amendments not linked to recent takedowns: MeitY Secy" — https://www.business-standard.com/industry/news/draft-it-rules-amendments-not-linked-to-recent-takedowns-meity-secy-126040800036_1.html — (Tier 4)
• [S9] The Hindu (Article — primary source excerpt) — "What has govt. laid down on AI labelling?" by Aroon Deep, Feb 15, 2026 — https://www.thehindu.com/todays-paper/2026-02-15/th_international/articleGJFFJCLBL-13512399.ece — (Tier 4)