Kamal Ghorui
NEW DELHI: Cracking down on deepfakes, the Centre has tightened India’s digital rules by mandating compulsory labelling, traceability and user declarations for AI-generated content. It has also sharply cut timelines to take down unlawful content to as little as three hours from 36 hours and placed direct compliance responsibility on social media platforms and their senior officers.For the first time, AI-generated material, including deepfake videos, synthetic audio and manipulated visuals, has been brought under a formal regulatory framework through amendments to the IT intermediary rules. The Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026 will come into force from Feb 20. The amended rules introduce a statutory definition of “synthetically generated information” (SGI).AI-made, edited posts must carry clear disclosureThe new rules impose mandatory obligations on intermediaries to identify, label and trace AI content. All AI-generated or AI-altered material must carry a clear and prominent disclosure visible to users. Platforms must also embed persistent metadata and unique identifiers to enable traceability of the content’s origin and the tools used to create it. Once applied, these disclosures cannot be altered, hidden or removed.The rules exclude routine edits such as colour correction, noise reduction, compression or translation, as long as they do not change the meaning. Clearly hypothetical or illustrative drafts are exempt.Large social media platforms face stricter compliance requirements. Before any content goes live, platforms must ensure users declare if it is AI-generated. Intermediaries must deploy automated tools to verify these declarations by analysing the content’s format, source and characteristics. If content is identified as synthetic, visible labelling becomes mandatory. Platforms that knowingly allow unlabelled AI-generated content to remain online will be treated as having failed their due-diligence obligations.The rules also compress response timelines. Certain lawful takedown orders must now be complied with within three hours, compared to 36 hours earlier. Other deadlines have been tightened, with a 15-day window reduced to 7 days and a 24-hour window cut to 12 hours. Platforms must acknowledge user grievances within two hours and resolve them within seven days.Oversight and enforcement will rest with the ministry of electronics and information technology, while users can appeal platform decisions to the grievance appellate committee.Misuse of SGI linked to child sexual abuse material, obscene content, false electronic records, impersonation using a real person’s identity or voice, or explosives-related material will attract action under multiple criminal laws.Platforms must also warn users at least once every three months about penalties for misuse of AI-generated content.
