Govt outlines regulatory steps against AI deepfakes

India’s legal framework against deepfakes: what happened?

A deepfake is AI-made audio, video or text that looks real. It can cheat people, spread lies or harm someone’s dignity. India has no single deepfake law. Instead, older laws on cheating, impersonation and privacy are applied, and the IT Rules were amended in 2026 to deal with AI content directly.

• IT Rules amendment of 10 Feb 2026: clear labels and traceable metadata for permitted AI-generated content.
• Removal of unlawful content after a government or court notice: 36 hours cut to 3 hours.
• Grievance redressal: 72 hours cut to 36 hours; for sensitive matters such as nudity or impersonation, 24 hours cut to 2 hours.
• Platforms must use technical measures to stop users from creating or sharing unlawful synthetic content.
• Advisories: 9 Feb 2026 (religious matters), 16 Mar 2026 (abusive or misleading synthetic content); SoP on Non-Consensual Intimate Imagery released 11 Nov 2025.

Which laws apply to a deepfake?

What extra duties do big platforms have?

• Significant Social Media Intermediaries (SSMIs) have 50 lakh or more registered users in India.
• Messaging SSMIs must help trace originators of serious content.
• They must use automated tools to detect unlawful content, publish compliance reports, appoint local officers and keep an Indian address.
• A platform that fails these duties loses its Section 79 safe harbour and can face action.

Who polices cybercrime under the Constitution?

• List II (State List), Entry 1: public order.
• List II (State List), Entry 2: police (including railway and village police), subject to Entry 2A of List I.
• MeitY notes that States and UTs are therefore primarily responsible for preventing, detecting, investigating and prosecuting cybercrime.

What can a victim do?

1. Complain to the platform’s Grievance Officer
2. If not resolved, appeal to a Grievance Appellate Committee (gac.gov.in)
3. Report on cybercrime.gov.in or call 1930
4. Police investigate; I4C and SAHYOG send removal notices
5. Platform must remove within the set timeline or lose safe harbour

Why does India’s legal framework against deepfakes matter?

What are the challenges in India’s legal framework against deepfakes?

• There is no single deepfake law; protection is spread across many provisions.
• A 3-hour takedown is demanding; platforms need good detection tools.
• Investigation rests with State police, whose cyber capacity varies.
• Labelling depends on platforms detecting AI content correctly.

What should be done next on India’s legal framework against deepfakes?

• Build detection tools such as Saakshya, AI Vishleshak and IIT Kharagpur’s voice deepfake system into wider use.
• Train State police in AI-related crime.
• Keep public awareness going through CERT-In advisories and Cyber Jagrookta Diwas.
• Review how the 2026 timelines work in practice.

What can Prelims ask on India’s legal framework against deepfakes?

• IT Rules amended 10 Feb 2026: labels and metadata for AI content; takedown 36 hours to 3 hours.
• SSMI: 50 lakh or more registered users in India.
• BNS Section 319: cheating by personation; Section 353: misinformation.
• SAHYOG portal: run by I4C for removal notices.
• India AI Governance Guidelines: 5 Nov 2025.
• Cyber helpline: 1930.

How to write about India’s legal framework against deepfakes in Mains

Practice question (15 marks, 250 words): Deepfakes pose new challenges to privacy, security and democratic discourse. Examine the adequacy of India’s legal and institutional framework to deal with them.

• Intro: define deepfakes, harms.
• Framework: IT Act, BNS, IT Rules 2021 and the 2026 amendment.
• Institutions: GACs, I4C, SAHYOG, CERT-In, State police.
• Gaps: no single law, capacity, detection.
• Way forward: detection tools, training, awareness.