India tightens AI rules for social media platforms

India's New AI Rules: What Social Media Platforms Must Do Now

India has introduced sweeping amendments to its Information Technology Rules, 2021, targeting AI-generated content on social media platforms. Effective from February 20, 2026, the rules mandate clear labeling of synthetic content, reduce takedown timelines for unlawful posts to three hours, and impose strict accountability on platforms like Instagram, YouTube, and Facebook. These measures aim to curb deepfakes, misinformation, and other harmful AI-generated material while balancing innovation with user protection.

Mandatory AI Labeling: Permanent Markers and Traceability

Platforms that allow users to create or share AI-generated content must now prominently label it. The label must be visible, unambiguous, and impossible to miss—no hidden tags or fine print. For visual content, the label must cover at least 10% of the total surface area; for audio, it should appear within the first 10% of the duration. Additionally, platforms must embed permanent unique metadata or identifiers to trace the content's origin, such as the system that created or distributed it. Once a label is applied, removal is prohibited. This ensures that AI-generated audio, video, or images are clearly distinguishable from authentic content, helping users identify deepfakes and synthetic media.

Three-Hour Takedown: Faster Action Against Unlawful Content

Perhaps the most significant change is the drastic reduction in the takedown timeline. Previously, platforms had 36 hours to remove content deemed illegal under Indian law. The new rules slash this window to just three hours after receiving a government notification. Authorities can order removal of content related to national security, public order, child sexual abuse, non-consensual intimate imagery, impersonation, forged documents, or explosives-related material. This puts immense pressure on platforms to swiftly detect and act on problematic posts, with non-compliance potentially leading to loss of safe harbor protection.

User Declarations and Automated Verification

Before any upload goes live, platforms must require users to declare whether the content is synthetically generated. They must also deploy automated tools to verify these declarations. If a declaration or technical check confirms AI generation, the platform must display a visible label. For large platforms with significant user bases, this obligation becomes even stricter. They are also required to send periodic reminders—at least once every three months—about the rules and consequences of violating them. This framework shifts responsibility to both users and intermediaries, ensuring that unlabeled synthetic content does not slip through the cracks.

Legal Liability and Safe Harbor Implications

Under the updated rules, if a platform knowingly permits or promotes unlabeled synthetic content, it is deemed to have failed due diligence. This could cost the platform its legal immunity (safe harbor) from third-party content liability. The rules also ban providing tools or services that help others tamper with these labels. Platforms that facilitate the creation or dissemination of AI content must now implement safeguards to prevent illegal synthetic media, including child sexual abuse material, obscene content, and deepfakes. Failure to comply can lead to account suspension, content removal, disclosure of user identity to victims, and mandatory reporting to law enforcement.

Definition of Synthetic Content and Exclusions

The rules formally define "synthetically generated information" as content artificially or algorithmically created, modified, or altered to appear authentic or true. This targets deepfakes and AI-generated impersonations. However, routine editing, accessibility features, academic and training materials, and good-faith technical changes that do not materially alter the content are excluded. This narrow definition aims to strike a balance—regulating harmful AI misuse while not hindering legitimate creative or educational uses. By focusing on content that mimics reality, the government addresses the most pressing concerns about AI deception and misinformation.

Global Implications and Platform Readiness

India's new rules are among the most stringent globally, setting a precedent for other nations grappling with AI regulation. With over 600 million internet users, India is a massive market for platforms like Meta, Google, and X. These companies already have some AI labeling practices but must now comply with India's specific requirements. The challenge lies in implementing tamper-proof labeling and three-hour takedowns at scale. While the intent is clear—protecting users from deepfakes and harmful content—experts warn that reliable labeling technology is still evolving. Platforms will need to invest in automated detection and metadata systems to avoid penalties. As the rules take effect, the world will watch how India shapes the future of AI governance in the digital ecosystem.