4. Regulatory Oversight of Artificial Intelligence: The IT Amendment Rules 2026

The Union Government has notified the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026, marking a decisive step in regulating the burgeoning field of Generative AI. Effective from February 20, 2026, these rules mandate the prominent labelling of \'synthetically generated information\' (SGI) and introduce ultra-short timelines for the removal of illegal content. This reform seeks to balance technological innovation with the need to protect citizens from digital harms like deepfakes, misinformation, and non-consensual intimate imagery. Summary of Key Developments • Mandatory AI Labelling: All \'photorealistic\' AI-generated content must be clearly and prominently labelled as synthetic. This ensures that users can immediately distinguish between real-world events and algorithmically created media. • Compressed Takedown Timelines: The window for removing unlawful content has been slashed from 36 hours to just 3 hours for court or government-ordered removals. For sensitive content like nonconsensual deepfakes or nudity, platforms must act within 2 hours. • Loss of Safe Harbour: Platforms that \'knowingly permit\' or fail to act against unlabelled synthetic content risk losing their \'Safe Harbour\' protection under Section 79 of the IT Act, 2000, potentially making them liable for user-posted content. • User Disclosure Obligations: Social media intermediaries must now require users to declare at the point of upload whether their content is AI-generated. Platforms are expected to use automated tools to verify these declarations. • Traceability and Metadata: Intermediaries are required to embed persistent metadata or technical provenance markers (unique identifiers) into AI content to enable its tracing back to the source platform. • Definition of Synthetic Media: The rules define Synthetically Generated Information (SGI) as audio, visual, or audio-visual info created or modified using computer resources that appears \'real, authentic, or true.\' Key Definitions • Synthetically Generated Information (SGI): Information created or altered algorithmically that is indistinguishable from real-world events or natural persons. • Safe Harbour: A legal provision (Section 79 of IT Act) that protects intermediaries (like X or YouTube) from being held liable for third-party data or content, provided they follow \'due diligence\' and government takedown orders. • Metadata/Provenance Markers: Digital \'watermarks\' or data embedded within a file that provide information about its origin, creation time, and the tools used for modification. Constitutional and Legal Provisions • Article 19(1)(a) & 19(2): While citizens have the right to free speech, the government can impose \'reasonable restrictions\' in the interests of public order, decency, and the sovereignty of India—grounds often cited for takedown orders.• Article 21 (Right to Privacy): The Supreme Court in the Puttaswamy judgment identified privacy as a fundamental right. Mandatory removal of non-consensual deepfakes is a direct protection of digital dignity under Article 21. • Section 79 of the IT Act, 2000: The primary statute governing intermediary liability. The 2026 Rules are a form of \'delegated legislation\' that specifies the \'due diligence\' platforms must perform to keep this immunity. • Information Technology Rules, 2021: The parent rules which these 2026 amendments update, establishing the framework for grievance officers and digital ethics. Conclusion The 2026 IT Amendments reflect a shift toward \'proactive accountability\' rather than reactive moderation. By mandating labels and near-instant takedowns, India is positioning itself as one of the first major economies to implement a \'trust-based\' digital architecture for the AI era. While the move strengthens digital safety, its success will depend on the technical capacity of platforms to distinguish between malicious deepfakes and harmless \'smartphone touch-ups,\' which have been explicitly exempted from these rules. UPSC Relevance • GS Paper II: Statutory, regulatory, and various quasi-judicial bodies; Government policies and interventions for development. • GS Paper III: Challenges to internal security through communication networks, role of media and social networking sites in internal security challenges; Basics of cyber security. • Mains/Ethics: Ethical concerns regarding Deepfakes, misinformation, and the \'Right to be Forgotten.\'