Creators and Small Businesses: AI Labelling Rules In India Has Set, and How to Follow Them

Sorry, This is too late… But I wanna share it with you guys, Only Because ? It’s Useful.

Walk through any market in a mid-sized Indian town this festival season and count the posters made with AI. The jewellery shop’s smiling family that does not exist. The coaching centre’s topper photo with slightly odd fingers. The restaurant’s biryani that looks better than any biryani has ever looked. A year ago, nobody asked whether these images needed a label. Today, the question is everywhere.

The reason is the set of AI labelling rules in India brought in this year through an amendment to the IT Rules, 2021. Most of what you will read about it is written for lawyers and for big platforms. Very little is written for the person who actually makes the reel, runs the shop’s Instagram page or designs the ad on Canva at night.

This article is for that person. I will explain in plain language what changed, what counts as synthetically generated information, who must do what under the IT rules amendment 2026, how the AI label on Instagram and other platforms fits in, what AI content disclosure means for ads and product images, and a one-page checklist you can follow from today. I will also be clear about what is still not settled, because pretending everything is clear would be the least honest thing I could do in a guide about honesty.

A small note before we start: I am a founder, not a lawyer. Treat this as a practical guide, and speak to a legal professional if your business depends heavily on AI content.

The Diwali Poster That Made Me Check My Own Posts

It was a weekday evening. I got down from the office bus one stop early to buy sweets for the weekend, and the sweet shop near our colony had a big new poster at the entrance. A family of four, all in silk, laughing around a plate of kaju katli. The light was perfect. The little girl had six fingers on one hand.

The shop owner saw me looking and said, proudly, that his nephew had made it “with AI” in ten minutes. He had printed twenty copies and posted it on Instagram too. I paid for half a kilo, said it looked nice, and walked home thinking about the four images I had used on Ethical Founder that month.

That night, instead of opening the laptop for client work, I opened my own posts. Two of them had AI images with no mention of AI anywhere. Nothing illegal had happened. But I knew the AI labelling rules India had notified in February, and I knew I had not read them properly. This article is the reading I should have done that month.

AI Labelling Rules IN India: What Changed on 20 February 2026

Let me start with the timeline of the AI labelling rules India now follows, because the dates matter.

According to an analysis by the law firm Hogan Lovells, the Ministry of Electronics and Information Technology (MeitY) notified amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 on 10 February 2026. The amendments came into force on 20 February 2026, giving platforms a ten-day window to get ready.

The three big changes, in one breath

The AI labelling rules India introduced through this amendment can be summed up in three parts:

1. A definition. The rules now define synthetically generated information, which covers AI-made or AI-altered audio, images and video that look real.
2. Labels and markers. Permitted synthetic content must be prominently labelled and carry permanent metadata or provenance markers.
3. Faster action. Platforms must act on government or court takedown orders within three hours, down from 36 hours.

Why this matters to a small creator

Strictly speaking, most of the duties in the IT rules amendment 2026 fall on platforms: Instagram, YouTube, Facebook, X, AI tool makers and other intermediaries. But platforms pass those duties down to users through upload forms, declarations, labels and removal of content. So the AI labelling rules India set reach every creator and every small business that posts AI content, even if the law itself is addressed to the platform.

Synthetically Generated Information: What Counts

The heart of the AI labelling rules India now follows is one term: synthetically generated information, often shortened to SGI.

The definition in plain words

As Hogan Lovells summarises it, synthetically generated information covers audio, visual or audio-visual content that is artificially or algorithmically created, generated, modified or altered using a computer resource, in a way that appears real, and portrays people or events as indistinguishable from real persons or real-world events.

Read that slowly. Three ideas are packed inside it:

• Made or changed by AI or software. Not a regular photo straight from a phone.
• Looks real. A cartoon that is obviously a cartoon is not trying to pass as real.
• Shows people or events as if they were real. A fake family, a fake event, a fake voice.

What is excluded

The same summary lists exclusions. Routine, good-faith editing and technical fixes, like formatting, colour adjustment, noise reduction and compression, are not treated as synthetically generated information. Documents, presentations, educational materials, drafts and templates that do not create false records are also excluded, as are tools used only for accessibility, clarity, translation or discoverability that do not change the substance.

Examples for small businesses

These examples are my reading of the definition, not legal advice. The safe rule is simple: if a reasonable viewer could believe the person, voice or event is real, and it is not, treat it as synthetically generated information and label it.

The middle cases

Some cases sit in between. A real product photo placed into an AI-generated kitchen background. A real person’s photo with an AI-changed outfit. A real video with AI-generated crowd noise. These are the cases where the AI labelling rules India set will be tested in practice. Until platforms and courts settle them, my own rule is: when in doubt, disclose. A small label costs nothing. A complaint, a takedown or a loss of trust costs a lot.

Who Must Do What Under the IT Rules Amendment 2026

The IT rules amendment 2026 splits duties across different players. Here is who does what, based on the Hogan Lovells summary.

All intermediaries that allow synthetic content

Platforms and tools that let users create or share synthetically generated information must make sure permitted SGI is prominently labelled, and that it carries permanent metadata or provenance markers. Those markers must include unique identifiers linking the content to the platform’s computer resource. Platforms must not let anyone modify, suppress or remove those markers.

Significant social media intermediaries

India treats platforms with more than 50 lakh (5 million) registered users in India as significant social media intermediaries. Under the AI labelling rules India now has, these large platforms must:

• Ask users to declare whether their content is synthetically generated information.
• Use technical measures to verify those declarations.
• Label confirmed SGI with a clear label or notice.

If a large platform knowingly permits or promotes non-compliant synthetic content, or fails to act on it, that is treated as a failure of due diligence.

Users and creators

The law speaks mainly to platforms, but platforms pass the job on to you through:

• A declaration question when you upload (“Is this AI-generated or altered?”)
• Labels applied to your post
• Removal or reduced reach for content that breaks the rules
• Warnings and account action for repeated problems

The amendment also says intermediaries must remind users at least once every three months about the consequences of unlawful activity. So expect regular notices from platforms about AI content.

The 3-Hour Takedown Rule

The most dramatic change in the AI labelling rules India package is speed. Intermediaries must now act on government or court orders, including takedown orders, within three hours of receiving them. The earlier window was 36 hours.

For a creator, this means content can disappear very quickly if it is reported through official channels. There may be little time to explain or edit before it is gone. The best protection is not to post content that is likely to attract an order in the first place: deepfakes of real people, misleading synthetic news, or anything that could be seen as impersonation.

AI Label on Instagram, YouTube and Facebook

Most small businesses in India meet the AI labelling rules India set through the platforms they already use. Here is how to think about the AI label on Instagram and the others. Platform screens change often, so look for these options in your own app when you post.

AI label on Instagram and Facebook

Meta’s apps show an AI label on posts that Meta detects or that users declare as AI-made or AI-edited. When you post on Instagram or Facebook, look for an AI label or AI info option in the advanced settings or the post options before sharing. If your image or video is realistic and AI-generated, switch it on. Meta also reads technical signals and metadata from some AI tools, so a label may appear even if you did not add it.

Two practical tips for the AI label on Instagram:

• Do not strip metadata on purpose. Some people screenshot AI images to remove the AI markers. Under the new rules, platforms are expected to preserve markers, and deliberately hiding them can look like bad faith.
• Add a caption line as well. A simple “Image made with AI” in the caption works on every platform and is understood by every customer.

YouTube

YouTube asks creators, during upload, whether the video contains realistic altered or synthetic content. If you answer yes, YouTube adds a label to the video. For small businesses making AI explainer videos or AI-voiced ads, answer honestly every time.

X, LinkedIn and others

Other platforms have their own labels or policies, and many are adding declaration options because of the AI labelling rules India now enforces on large platforms. Where no label option exists, use the caption or the first line of the post.

If you use AI image tools heavily, my comparison of Nano Banana Pro vs ChatGPT image accuracy shows how realistic these tools have become, which is exactly why the AI label on Instagram and elsewhere now matters.

AI Content Disclosure for Ads and Sellers

Ads and product listings are where AI content disclosure matters most for a business, because money changes hands based on what the customer sees.

Ads

If you run ads with AI-generated people, AI voices or AI scenes that look real, disclose it. Ad platforms review creatives, and the AI labelling rules India set make undisclosed realistic synthetic content riskier. A small “AI-generated image” line on the creative or in the ad text is enough for most small businesses. If you are automating ad creation, our Facebook ad automation tool guide shows where to add that line in a workflow so it is never forgotten.

Product images

For products, honesty protects you twice: from the platform and from returns. If the product itself is shown accurately and only the background or setting is AI-made, many sellers add a short note like “Lifestyle image created with AI. Product shown is actual.” If the product’s colour, size or design is changed by AI, do not use the image at all. That is not a labelling problem, it is a misleading listing.

Influencer and creator posts

If you pay a creator to promote your product and they use AI-generated scenes or an AI version of themselves, ask them to disclose it. Your brand carries the reputation either way.

Testimonials and reviews

Never create AI customers, AI testimonials or AI review videos and present them as real. That is not AI content disclosure, it is a fake endorsement, and it can break consumer protection rules, platform policies and the AI labelling rules India now applies, all at once.

What Is Clearly Not Allowed

Some content was a problem long before the AI labelling rules India introduced in the IT rules amendment 2026, and the new rules make enforcement faster. Stay far away from:

• Deepfakes of real people, especially public figures, politicians, celebrities or private individuals, without consent.
• Non-consensual intimate content of any kind. This is illegal and harmful, and platforms act on it aggressively.
• Fake documents and records, like AI-made certificates, bills, ID cards or official letters.
• Synthetic news or fake events that could mislead people, like an AI video of a riot, a disaster or a government announcement.
• Impersonation of a brand, a government office or a person.
• Content involving children that is sexual, exploitative or harmful in any way.

The Hogan Lovells summary notes that non-compliance can lead to suspension or termination of access, content removal and potential penalties under the IT Act and other laws. For a small business, losing an Instagram or YouTube account can mean losing years of customers overnight. The AI labelling rules India set are not worth testing at the edges.

AI Labelling Rules India Checklist for Creators and Shops

Here is the checklist I now follow under the AI labelling rules India set, for every post on my own sites and social accounts. Print it, stick it near your desk, or save it in your notes.

Before posting

1. Is any person, voice or event in this content AI-made or AI-altered and realistic?
2. If yes, is it a real person who has given consent? If not, do not post.
3. Does the AI change the product, price, result or claim? If yes, do not post.
4. Keep the original file and the prompt or source, in a folder with the date.

While posting

1. Answer the platform’s AI declaration question honestly.
2. Turn on the AI label on Instagram, Facebook or YouTube where available.
3. Add a plain caption line: “Image made with AI” or “AI voice-over”.
4. Do not screenshot or strip the file to remove AI markers.

After posting

1. If someone asks whether it is AI, answer plainly.
2. If the platform labels or removes it, read the notice, fix the issue, and do not repost the same file.

For your team

If someone else designs or posts for you, share this checklist with them. Under the AI labelling rules India now has, the account owner carries the consequence, not the designer.

If you are automating your content, as in our X content creation workflow or podcast clip automation, add the AI content disclosure line as a fixed step in the workflow, so the label is added every time without anyone remembering.

What Still Is Not Clear in the AI Labelling Rules India Set

I promised to be honest about the open questions. Here they are, as of the time of writing.

The exact look of a “prominent” label. The rules require prominent labelling, but small creators do not yet have one standard design to copy. Platforms are deciding how their labels look.

How verification will work. Large platforms must verify user declarations with technical measures. How accurate those checks will be, and how disputes will be handled, will become clearer with time.

Middle cases. Real photos with AI backgrounds, AI-enhanced faces, AI-cleaned audio. Where routine editing ends and synthetically generated information begins will be tested case by case.

Enforcement on small accounts. Most early enforcement is likely to focus on harmful deepfakes and large platforms. But small accounts can still lose posts or reach through platform systems.

My advice for all of these is the same: label generously, keep your originals, and avoid anything that could mislead a customer. The AI labelling rules India set reward people who are open about how they make content.

If you follow AI news to stay ahead of changes like this, my notes on the GPT-5.1 update, the JEV AI model and the Qwen Image 2.1 lawsuit the AI agents market in 2026 and my Grok AI agents guide for solo founders show how fast the ground is moving. Small businesses that keep their content honest will be the ones least affected each time it moves again.

AI Labelling Rules India by Type of Small Business

The AI labelling rules India introduced touch every business differently, depending on how it uses images, voice and video. Here is how I would apply them to the kinds of businesses most of my readers run.

Kirana stores, sweet shops and local retail

Most local shops use AI for festival posters and WhatsApp Status images. If the poster shows a realistic family, crowd or person, add a small “AI image” line on the poster itself and in the caption. If the poster only uses AI for decorations, diyas, patterns or text, there is usually nothing realistic to label. Under the AI labelling rules India now follows, the real question is always whether someone could mistake a made-up person or event for a real one.

Coaching centres and schools

This is a sensitive area. Never use AI to create “toppers”, “students” or “results” that look real. That is not a labelling question, it is misleading advertising aimed at parents. Use real photos with permission, or clearly illustrated characters. If you are opening a centre in a smaller place, my list of business ideas for small towns explains why trust decides everything there. For AI explainer videos and AI voice lessons, a label in the video and the description is enough.

Restaurants and cloud kitchens

AI food photos are tempting because they always look perfect. But a customer who orders the biryani from an AI picture and receives something different will feel cheated. If you use AI food images, label them and keep them close to the real dish. Better still, use real photos for the menu and AI only for festival greetings.

Real estate and interiors

AI-generated room designs and building views are common. Label them as “AI concept view” or “representation”. Never present an AI image as the actual flat, view or completed project. The AI labelling rules India set, plus existing consumer and real estate rules, make this an area where honesty matters more than ever.

D2C brands and online sellers

Your main product photo should be the real product. AI can help with lifestyle scenes, backgrounds and models, but the product’s colour, size and design must stay true. Add AI content disclosure for realistic AI models and scenes. My upcoming guide on AI product images for marketplaces covers this in detail. And if you are still deciding how to promote a new brand, the day a small business owner stops doing everything is a good read on handing content work to someone else without losing control.

Service businesses and freelancers

If you use an AI avatar or AI voice to present your services, say so in the video and the caption. Clients who later meet you and realise the “person” in your ads was synthetic will lose trust quickly, even if you followed every rule.

Keeping Proof of Your Original Content

One part of the AI labelling rules India set gets less attention: what happens when your real content is mistaken for AI, or someone claims your AI content is misleading. In both cases, proof helps.

What to keep

• Original camera files for real photos and videos, with the date.
• The AI tool, prompt and date for AI images and videos.
• The final file you posted, with the label visible.
• Consent messages from any real person who appears in your content.

Where to keep it

A simple folder system works: one folder per month, one sub-folder per post. Cloud storage is fine. The point is that if a platform labels or removes your post, or a customer asks, you can show exactly how the content was made within minutes.

Why it matters now

Platforms are expected to verify declarations using technical measures. No automatic system is perfect, and real photos are sometimes flagged as AI. Having your original files ready makes it far easier to appeal. Under the AI labelling rules India now follows, the creator who can show the original is always in a stronger position than the one who cannot.

Why This Is Good News for Honest Small Businesses

It is easy to read the AI labelling rules India introduced as one more burden. I see it differently.

For years, small businesses have competed with fake reviews, fake followers and too-good-to-be-true pictures. The new rules push the whole internet a little closer to honesty. A shop that labels its AI poster and shows its real products builds more trust than one pretending a fake family is its customer.

Trust has always been the real currency of a small business. My guide on word of mouth vs ads for a first-year business makes the same point from a different angle: customers remember who was straight with them. And for founders in smaller places, the business ideas for small towns that work best are almost always the ones built on visible honesty.

So use AI. Use it for posters, product scenes, explainer videos and voice-overs. Just say so. That one line of AI content disclosure costs nothing and protects everything.

Questions Creators Keep Asking About AI Labels