How Fake News and Disinformation Threaten Internal Security and Public Order — and the Salient Features of the IT Rules Amendments
A complete, examiner-standard 150-word model answer for the UPSC Mains 2026 GS Paper 3 question on disinformation — with a salient-features table, three pointer groups, static core content, and the amendment the question is built on: the IT Amendment Rules notified 10 February 2026, in force from 20 February 2026.
The examiner has dated this question precisely. "Amendments in respect of the IT Rules 2021" in an August 2026 paper means the February 2026 synthetic-content amendment. A candidate who described the original 2021 Rules and stopped has answered the question as it stood three years ago.
Explain how fake news and disinformation pose threat to Internal Security and Public Order in Indian context. In this regard, discuss salient features of amendments in respect of Information Technology (Intermediatory Guidelines and Digital Media Ethics Code) Rules 2021. (Answer in 150 words)
Model Answer
Introduction
Fake news threatens public order less through what it asserts than through the speed at which it mobilises. Order collapses in the interval between a rumour's circulation and its verification — an interval that generative AI has now compressed to almost nothing.
Body
Part I — The Threat
- Communal mobilisation — doctored or decontextualised imagery circulating on closed messaging groups has repeatedly preceded local violence, converting a rumour into a law-and-order emergency within hours and forcing preventive internet shutdowns.
- Mob violence and panic — forwarded rumours about child-lifting or cattle theft have produced lynchings; false claims of targeted attacks have triggered the exodus of migrant workers from host States.
- Deepfake impersonation — synthetic audio or video of officials, security personnel or public figures can manufacture an announcement that never happened, with consequences for both public order and financial fraud.
- External information operations — coordinated campaigns during border tensions or counter-terror operations aim to shape domestic perception and erode confidence in official information, making disinformation a hybrid-warfare instrument rather than a purely domestic policing problem.
- Disaster and election contexts — false information during relief operations diverts scarce resources; during elections it corrodes the legitimacy of the outcome itself.
Part II — Salient Features of the Amendments
| Feature | What the 2026 Amendment Introduces |
|---|---|
| New definition | "Synthetically generated information" (SGI) — audio, visual or audio-visual information artificially or algorithmically created or altered so as to appear real and depict a person or event deceptively. Technology-neutral: it turns on the content, not the tool. |
| SGI as "information" | SGI is expressly brought within the unlawful-content provisions of the 2021 Rules, closing the argument that deepfakes occupied a regulatory gap. |
| User declaration | Significant social media intermediaries must obtain a declaration from users on whether uploaded content is synthetic, before publication. |
| Labelling and provenance | Lawful SGI must be prominently and visibly labelled, with provenance metadata embedded where technically feasible; labels and identifiers may not be removed or obscured. |
| Technical measures | Platforms whose tools can generate SGI must deploy automated means to prevent creation of unlawful synthetic content — CSAM, non-consensual intimate imagery, false documents or records, and deceptive depictions of persons or events. |
| Compressed timelines | Takedown on a court order or authorised government notice compressed to roughly three hours, from the earlier thirty-six-hour window, with shorter periods for the gravest categories. |
| Authorising officers | Reasoned intimations may now be issued by one or more officers not below the rank of Deputy Inspector General, to sustain round-the-clock capacity. |
- Process — MeitY released draft rules for consultation in October 2025, notified the final Amendment Rules on 10 February 2026, and brought them into force on 20 February 2026, a ten-day interval. The final version narrowed the SGI definition to exclude routine and good-faith uses following stakeholder feedback.
- Compliance is tied to safe harbour — the operative sanction is not a fine but loss of the statutory immunity intermediaries have enjoyed under Section 79 of the IT Act, which is why platforms have moved quickly on labelling.
- The case for — the harm from synthetic impersonation and coordinated falsehood is immediate and often irreversible; a thirty-six-hour window is meaningless when a deepfake achieves saturation in minutes. Labelling gives users the information needed to judge authenticity for themselves.
- The case against — critics argue that mandatory labelling of all synthetic content amounts to compelled speech; that conditioning safe harbour on proactive monitoring sits uneasily with the reasoning in Shreya Singhal v. Union of India (2015); that the SGI definition is broad; and that watermarking and AI-detection tools perform poorly, particularly across Indian languages.
- Comparative note — the EU AI Act tiers obligations by risk and treats benign synthetic content differently from harmful content; India's approach applies labelling more uniformly.
- Practical risk — a three-hour clock creates an incentive to over-remove rather than adjudicate, since the cost of wrongful takedown falls on the user and the cost of delay falls on the platform.
Conclusion
The amendment closes a real regulatory gap: before it, deepfakes were governed by rules drafted for a pre-generative-AI internet. Whether it succeeds depends less on the takedown clock than on capacity — reliable multilingual detection, trained investigators below the DIG rank, an independent check on removal orders, and digital literacy that reduces reliance on takedowns altogether. Speed of removal treats the symptom; verification capacity treats the disease.
Legal framework: Information Technology Act, 2000 — Section 69A on blocking, Section 79 safe harbour, Section 87 rule-making power. Shreya Singhal v. Union of India (2015) struck down Section 66A and read down Sections 69A and 79, and remains the governing precedent on online speech. Revise the original IT Rules 2021 — the three-tier grievance redressal architecture, the Grievance Appellate Committees added in 2022, the additional obligations on Significant Social Media Intermediaries including the Chief Compliance Officer, Nodal Contact Person and Resident Grievance Officer, traceability of the first originator, and Part III on digital news and OTT with its Code of Ethics. Also revise the Bharatiya Nyaya Sanhita provisions on promoting enmity and on circulating false information, and the Digital Personal Data Protection Act, 2023.
Institutional and doctrinal: Indian Computer Emergency Response Team (CERT-In), Indian Cyber Crime Coordination Centre (I4C), the National Cyber Crime Reporting Portal, the PIB Fact Check Unit, and the National Cyber Security Policy. On the concepts, distinguish misinformation (false, shared without intent to harm), disinformation (false, deliberately deployed) and malinformation (true but weaponised through selective disclosure), and know the reasonable-restrictions test under Article 19(2) — public order, decency, incitement to an offence — which is the constitutional lens for every question in this area.
Answer Writing Tips for This Question
- The amendment is the point of the question. "Salient features of amendments" carries roughly half the marks. Give it a table and its own section — describing the 2021 Rules and adding one line on deepfakes will read as though you stopped following the subject in 2023.
- Carry the dates and the term of art. Notified 10 February 2026, in force 20 February 2026, following an October 2025 draft consultation; the operative concept is "synthetically generated information", not "deepfake regulation". Using the statutory vocabulary signals that you read the notification.
- Distinguish internal security from public order. The stem names both and they are not synonyms. Public order is the immediate law-and-order consequence — riot, lynching, panic. Internal security is the strategic dimension — hybrid warfare, foreign information operations, erosion of institutional trust. Separating them earns marks that merging them forfeits.
- Anchor the constitutional test. Shreya Singhal (2015) and the Article 19(2) reasonable-restrictions framework are what convert a description of rules into legal analysis. Almost no candidate brings the safe-harbour logic into this answer.
- Present both assessments as positions, not verdicts. The government's rationale — the speed and scale of synthetic harm — and the constitutional critique of compelled speech and prior restraint are both worth stating fairly. This is a live policy dispute; an answer that argues one side reads as advocacy and scores lower for it.
- Do not name individuals, platforms or parties as offenders. Describe categories of harm and the legal response. Naming a specific viral incident, company or political actor as the wrongdoer adds no marks and introduces risk on a sensitive topic.
