AI and Deepfake Content Labelling Rules in India 2026

Yes, you can post AI generated or deepfake videos in India, but from 20 February 2026 you must clearly label them as synthetic. If a deepfake targets you, you can report it to the platform and have unlawful or intimate-image content removed quickly. This guide explains the new labelling duty under the IT Intermediary Amendment Rules 2026 and the exact steps to get a fake video pulled down.

The Ministry of Electronics and Information Technology notified the Information Technology, Intermediary Guidelines and Digital Media Ethics Code, Amendment Rules, 2026 through Gazette notification G.S.R. 120 E dated 10 February 2026. The rules came into force on 20 February 2026 and are not stayed. They create a clear legal duty to label AI generated content and a faster system to remove harmful deepfakes.

Use this simple flow before you post anything made or changed with AI.

1. Did you create or alter this image, audio or video using AI? If yes, it is treated as synthetically generated information. This covers deepfakes, voice clones and face swaps.
2. Must you label it? Yes. Visual content must carry a prominent, easily noticeable label that a normal viewer can perceive. Audio must carry a clearly prefixed spoken disclosure that it is AI generated.
3. What happens if you do not? You lose the protection given to ordinary content. The platform can be told to remove your post, and unlabelled or deceptive synthetic media that breaks the law can be taken down. Impersonation and fake intimate images carry the heaviest consequences.

The aim is simple. People should be able to tell when something is real and when a machine made it. The law does not ban AI content. It bans hiding that it is AI content.

Under the amended rules, synthetically generated information means content that is artificially or algorithmically created, generated, modified or altered, and that appears to be real, authentic or true. The test is whether an ordinary person could mistake it for a real person or a real event.

So the following are covered:

• A video that puts a real person's face on someone else's body, a face swap.
• A voice clone that makes a leader or a relative appear to say something they never said.
• An AI image that shows an event that never happened in a realistic way.
• AI altered photos and synthetic audio used to deceive or defame.

A purely cartoonish or obviously artificial creation that no one would take as real sits at the edge of this definition. When in doubt, label it. Labelling is cheap and it protects you.

The rules use a quality standard, not a fixed size. An earlier draft had suggested that a label must cover ten percent of the screen, but that figure was dropped from the final notified rules. What the law now requires is this.

• For visual content, the label must be prominent, easily noticeable and adequately perceivable. It must not be hidden in a corner where viewers will miss it.
• For audio content, there must be a clearly prefixed audio disclosure stating that the sound is synthetically generated.
• Provenance, larger platforms are also expected to keep permanent metadata or technical markers, including a unique identifier, so synthetic content can be traced.

If you are a creator, place a visible on screen tag such as “AI generated” and add a spoken line at the start of any synthetic audio. Keep it through edits and re-uploads.

A Significant Social Media Intermediary is a platform with more than fifty lakh registered users in India. These large platforms now carry extra duties.

• They must ask you to declare, at the time of upload, whether your content is synthetically generated.
• They must deploy reasonable technical tools to check whether that declaration is accurate.
• They must label content their tools detect as synthetic, even where a user did not declare it.

In short, you should expect an upload screen that asks “Is this AI generated?” The honest answer protects you. A false declaration can expose you to action by the platform and under law.

A fake video, cloned voice or morphed image of you is exactly what these rules are built to tackle. Here is the practical route.

1. Step 1, capture proof. Save the link, take screenshots and note the date and time. Do not only rely on the post staying up.
2. Step 2, use the platform's report tool. Pick the closest category, such as impersonation, non consensual intimate image, or misleading synthetic media. File a complaint with the platform's Grievance Officer, whose contact must be published on the site or app.
3. Step 3, know the clocks. A non consensual intimate image complaint must be acted on within two hours of receipt. A complaint about other prohibited content must be resolved within thirty six hours. Ordinary grievances must be resolved within seven days. When a platform has actual knowledge through a court order or a government direction, it must remove unlawful content within three hours.
4. Step 4, escalate. If the platform does not act or you are unhappy with its reply, appeal to the Grievance Appellate Committee within thirty days of the platform's decision.
5. Step 5, go to the authorities. For serious deepfakes, including fake intimate images, file a police complaint, and report it on the National Cyber Crime Reporting Portal at https://cybercrime.gov.in.

The three hour clock explained. The three hour window is the time a platform gets to remove unlawful content once it has actual knowledge through a court order or a government direction. It is not automatically started by your private complaint. Your fastest direct lever as a citizen is the two hour rule for non consensual intimate images and the platform grievance timelines above.

Government bodies, including MeitY, the police cyber cells and grievance authorities, hold records on how complaints are handled. If your complaint about a deepfake is ignored, you can use the RTI Act 2005 to ask a public authority what action it took, what timelines it followed and why. An RTI under Section 6 of the Act can surface the paper trail that a platform or police station would rather not share. You can prepare a clean request using our AI RTI Drafter and, if you get no reply, build your appeal with the First Appeal Builder. For the full method, see The RTI Playbook.

Yes, posting AI or deepfake content is legal, but from 20 February 2026 you must label it as synthetically generated. Content that defames, impersonates, or shows non consensual intimate images is unlawful and can be removed and prosecuted, label or no label.

Yes. Visual synthetic content must carry a prominent, easily noticeable label, and synthetic audio must carry a clearly prefixed spoken disclosure. Large platforms must also ask uploaders to declare AI content and verify those declarations.

No. That figure appeared in an earlier draft and was dropped from the final notified rules. The law now requires a label that is prominent, easily noticeable and adequately perceivable, without fixing an exact size.

A non consensual intimate image complaint must be acted on within two hours. Other prohibited content must be resolved within thirty six hours, and ordinary complaints within seven days. A court or government order obliges removal of unlawful content within three hours.

It is a platform with more than fifty lakh registered users in India. Such platforms carry extra duties, including collecting user declarations about AI content and using technical tools to verify and label synthetic media.

First file with the platform's Grievance Officer. If unsatisfied, appeal to the Grievance Appellate Committee within thirty days. For criminal deepfakes, report to the police and at https://cybercrime.gov.in.

• MeitY Gazette notification G.S.R. 120 E dated 10 February 2026, in force 20 February 2026.
• IndiaLaw, IT Rules 2026 on synthetic media, definition and labelling analysis.
• Khaitan and Co, MeitY notifies the IT Amendment Rules 2026, labelling standard and takedown timelines.
• Argus Partners, overview of the IT Intermediary Amendment Rules 2026, grievance and removal timelines.

• The RTI Act 2005, full text
• AI RTI Drafter
• First Appeal Builder

This guide is general information, not legal advice. Reviewed by the editorial team led by Dr. Shrawan Kumar Pathak.