AI-Made Ads Now Need Labels in India: What Brands Must Do

If you run ads in India that were made or meaningfully edited with AI, you now need to say so. India's amended IT Rules, notified in February 2026, require synthetically generated content to be clearly labelled, and the ad platforms have built the tooling to match. Google added an AI label setting for ad creatives in July 2026, and its own policy page names India among the regions whose rules require disclosure.

This is not a future problem. It is live, it applies to ordinary brand advertising, and most Indian businesses running AI creative have not noticed.

The short version

• India's IT Amendment Rules, 2026 introduced a formal category called Synthetically Generated Information and require it to be labelled prominently.
• Google and Meta have both shipped disclosure tools for advertisers, so the practical burden now sits with you.
• Light touch edits like cropping or colour correction are generally not the concern. Synthetic people, scenes, voices and video are.
• The cheapest fix is a workflow habit: record what was AI-made at the point of production, not months later.

What counts as AI-generated content?

India's rules define Synthetically Generated Information as audio, visual or audio-visual content that is created or altered by computational means so that it appears authentic or real. That definition is broad on purpose, and it is worth reading carefully, because it covers far more than deepfakes.

In practical advertising terms, you should assume you are in scope if your ad includes:

• A person, product or environment generated by an AI image or video model
• A synthetic or cloned voice in a video or audio ad
• A real scene altered enough that a viewer would not know what was changed
• AI generated video used as the main creative rather than a small element

You are generally not in scope for routine post-production: cropping, resizing, colour grading, retouching a blemish, or cleaning up audio. The line is roughly whether an ordinary viewer would be misled about what they are looking at.

If you are honestly unsure where a piece of work falls, treat that uncertainty as your answer and disclose. Disclosure costs you very little. Getting it wrong costs more.

What do Google and Meta actually require?

The platforms have moved faster than most advertisers realise.

Platform	What changed	What you do
Google Ads	July 2026, added an AI label setting, and confirmed labels placed inside creatives will not breach its text overlay rules	Use the label setting, or build the disclosure into the creative itself
Meta	Reported since early 2026, requires advertisers to declare AI generated or AI modified creative	Apply the AI content label in Ads Manager when you submit the ad

Two details matter more than the rest.

Google may label its own AI output automatically, but not everyone else's. Assets generated with Google's own tools can be labelled for you. Work produced in third party tools, which is most studio work, relies on you declaring it. Do not assume silence means you are covered.

Google says the label is not the same as compliance. Its policy page is explicit that using the AI label setting does not guarantee you meet any particular regulation. The platform setting is a tool, not a legal shield. Your obligation under Indian law sits with you.

Reported enforcement on Meta follows the usual pattern of rejected ads and policy strikes accumulating against an account. Treat that as a strong reason to get the habit right rather than a reason to panic.

Does this apply to a small brand, or only to big advertisers?

The platform rules apply to anyone buying ads, at any budget. A small D2C brand in Ahmedabad running a Diwali campaign with an AI generated hero visual is in exactly the same position as a national advertiser.

The Indian rules place heavy obligations on the large platforms, including verifying user declarations rather than taking them at face value. The practical effect flows downhill to you: the platform is required to ask, so you will be asked, and your answer needs to be correct.

Will labelling hurt your ad performance?

There is no reliable public data yet showing that an AI label meaningfully depresses ad performance in India, so be sceptical of anyone who tells you confidently either way. What we can say is more useful.

Audiences have become good at spotting synthetic work on their own. An unlabelled AI ad that viewers recognise as AI reads as a brand trying to hide something, which is worse for trust than the label ever was. Declared AI work, made well, does not carry that problem.

The performance risk worth worrying about is not the label. It is bad AI creative: uncanny faces, physically impossible product details, generic stock feeling scenes. That was already costing you money before any of this changed.

What to do this week

1. Audit what is running now. List every live ad and mark which creative was AI generated or significantly AI altered.
2. Apply the platform labels to anything in scope, using Google's AI label setting and Meta's AI content label.
3. Add one field to your production checklist. Every asset gets recorded as AI generated, AI assisted, or neither, at the moment it is made. Reconstructing this later is painful and unreliable.
4. Brief your studio or freelancers. Ask them to declare AI usage on delivery. If a partner cannot tell you how a frame was produced, that is a bigger problem than the label.
5. Keep provenance where you can. The rules encourage embedded metadata that traces origin. Preserving it costs nothing and it is the direction the whole industry is heading.
6. Check the specifics with your counsel if you advertise at scale, in regulated categories, or on political and social issues, where the strictest rules apply.

This is a practical summary for marketers, not legal advice.

The honest take

We build AI video and AI artwork every day, so we will say the quiet part plainly: labelling is not a threat to studios doing this well. It is a threat to work that was hoping nobody would notice.

If your AI creative is good, saying it is AI costs you nothing. If your process cannot even tell you which assets were synthetic, that is worth fixing regardless of what any regulator says. We wrote more about building responsibly with AI in our guide to AI agents for Indian businesses.

If you want AI creative that holds up commercially and is clean on disclosure, that is exactly what our AI and film teams do. Tell us what you are producing this season and we will make sure it ships correctly. Get in touch.