IT Rules 2026: India Mandates Rapid Deepfake Takedowns and AI Content Labelling
Updated: Oct 15
India has formally moved to regulate deepfakes and AI-generated content through the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026. These amendments introduce a structured regime for synthetically generated information, imposing two-hour and three-hour takedown deadlines for different categories of harmful AI content and new labelling, traceability, and due diligence requirements for intermediaries. For social media platforms, AI companies, content creators, and digital publishers operating in India, the new rules create binding obligations with real enforcement consequences.
The Accelerated Takedown Mandate for Harmful AI Content
The most striking provision is the requirement that intermediaries remove certain categories of content within two hours of receiving a complaint, with a three-hour deadline for other unlawful content. The shorter timeline sits in rule 3(2)(b) and is confined to content which exposes the private area of the complainant, shows that person in full or partial nudity or in any sexual act or conduct, or is in the nature of impersonation in an electronic form, including artificially morphed images of that person. It runs from receipt of a complaint by the individual concerned or a person acting on their behalf. Nothing in the Rules attaches a two-hour window to defamatory impersonation, to impersonation likely to incite violence, or to the impersonation of government officials or institutions. The three-hour window in rule 3(1)(d) replaces a previous period of thirty-six hours, and it runs not from a complaint but from actual knowledge under Section 79(3)(b) of the Information Technology Act, 2000, which arises only from an order of a court of competent jurisdiction or a reasoned written intimation from an authorised officer. Intermediaries that fail to meet this deadline risk losing their safe harbour protection under Section 79 of the Information Technology Act, 2000, exposing them to direct liability for the content hosted on their platforms.
AI Content Labelling and Traceability Obligations
The amendments require that all synthetically generated content distributed through intermediary platforms carry a visible label or watermark identifying it as AI-generated. This labelling requirement applies to text, images, audio, and video content produced using generative AI tools. The rules place the primary labelling obligation on the creator or first uploader of the content, but intermediaries are required to develop and deploy technical measures to detect unlabelled synthetic content and apply labels where the original creator has failed to do so. On the traceability front, intermediaries must maintain records sufficient to identify the first originator of flagged synthetic content, building on the existing traceability requirements under the 2021 IT Rules. These provisions are designed to create an accountability chain from the creation of deepfake content to its distribution, making it possible for law enforcement to trace harmful synthetic material back to its source.
Due Diligence Requirements for AI Tool Providers
The rules also impose due diligence obligations on providers of AI tools and platforms that can be used to generate synthetic content. These providers must implement safeguards to prevent their tools from being used to create content that violates Indian law, including non-consensual intimate imagery, child exploitation material, and content designed to mislead the public on matters of public interest. AI tool providers must also maintain usage logs and make them available to authorised government agencies on request. While the rules do not ban generative AI tools outright, they create a compliance framework that requires AI companies to take proactive steps to prevent misuse of their technology within the Indian jurisdiction.
Practical Takeaways for Platforms and Content Creators
Social media platforms and significant social media intermediaries must upgrade their content moderation systems to act on a complaint engaging rule 3(2)(b) within two hours, and on actual knowledge engaging rule 3(1)(d) within three hours. This likely requires investment in AI-based detection tools and round-the-clock moderation capabilities. Content creators using generative AI tools should ensure that all synthetic content they publish carries appropriate labels, as failure to do so may result in the content being flagged or removed by platforms. AI companies offering generative tools in India should review their terms of service, implement output labelling by default, and establish cooperation mechanisms with Indian law enforcement. Digital publishers should train their teams to identify and report deepfake content that appears on their platforms or references their brand. The IT Rules 2026 amendments represent a significant regulatory escalation in India's approach to AI-generated content, and compliance readiness should be treated as a priority.
Comments