India Implements AI Content Rules: Deepfake Images Must Be Removed Within 3 Hours

New Delhi, February 21, 2026: In a significant step to curb the misuse of AI, the Indian government has made it mandatory to label all AI-generated photos, videos, and audio. Social media platforms are now required to remove any objectionable content within three hours of receiving a complaint. The new rules came into effect on Friday.

Notification and Implementation

The official notification was issued on February 10, 2026. IT Minister Ashwini Vaishnaw said several countries have praised India’s decision to enforce clear labeling of AI-generated content and are considering similar measures.

PM Modi’s Suggestion

A day before the rules were implemented, Prime Minister Narendra Modi had suggested at the AI Summit that digital content should carry labels similar to “nutrition labels” on food products. This would help citizens distinguish between genuine and AI-generated content.

Key Provisions of the New Rules

• Mandatory AI Labels: Every AI-generated photo, video, or audio must carry a visible label indicating it is AI-generated, similar to “vegetarian” or “non-vegetarian” labels on food packages.
• Technical Marker: While not visible on screen, the metadata embedded in the file will act as a technical marker, enabling authorities to trace misuse.
• Prevention of Tampering: Platforms must prevent users from removing or editing labels, watermarks, or technical markers from AI-generated content.
• Serious Offenses: AI misuse for child pornography, obscenity, fraud, weapons-related information, or impersonation will be considered a grave crime.

Addressing Misuse at Expos

Vaishnaw referenced controversies at recent AI events, including a demo by Galgotia University claiming a Chinese robotic dog as their invention. The university was removed from the expo after the misleading video went viral. The Minister emphasized that the government would take immediate action against attempts to undermine legitimate AI initiatives.