ASCI’s new AI rules:Do I have to label AI in my ads?
Sometimes, yes. If AI shows a person, a voice, a place or a result that could change what a buyer decides, label it. If it just fixed the lighting or wrote your caption, you don’t need to. And some things are not allowed even with a label. Here is how to tell, with 12 everyday examples.
By Madhur Tyagi
· 10 min read · Facts checked 3 Nov 2026
Quick answer
Do I have to label AI in my ads in India? Yes, when AI could change what a buyer decides and hiding it could mislead them: AI influencers, a cloned face or voice, made up settings, demos of things that don’t exist. ASCI’s guidelines, released 29 September 2026, take effect about three months later.
In short, for busy readers and AI answers
- Who made the rules?
- The Advertising Standards Council of India (ASCI), released in Mumbai on 29 Sep 2026.
- When do they start?
- Three months from publication, so around the end of December 2026.
- Must label
- AI influencers, a real person’s cloned likeness or voice, fabricated events or settings, demos of products that don’t exist, key product sounds, paid AI suggestions (“Sponsored by [Brand]”).
- No label needed
- Colour and light fixes, decorative elements, obvious fantasy, AI written copy, subtitles and translations.
- Not allowed, even with a label
- Fake testimonials, exaggerated results, fake places shown as real, deepfakes or likeness without consent, unauthorised copyrighted work.
Key takeaways
- ASCI’s test is about the buyer, not the tool. Label AI when it could change a decision and hiding it could mislead.
- Your own AI clone selling your course needs a label, even with your consent.
- AI written captions, subtitles and colour fixes need no label.
- Fake testimonials and fake results are banned. A label does not make them okay.
- Suggested labels: “Audio/Video created using AI” or “Audio/Video enhanced using AI”.
- Save the Jaggery Test: Choice, Real, Yes.
Written by Madhur Tyagi, founder of The Human Algorithm. 8+ years, 25+ brands, since 2017.
Your audience will forgive you for using AI.
They won’t forgive you for hiding it.
There is a sweet shop in north Kolkata that started printing four words on its boxes: “Made with jaggery, not sugar.”
You would think saying “this is different” scares buyers. It did the opposite. People trusted the box. They bought more boxes. The label became part of the reason to buy.
That is exactly how I want coaches and creators to think about AI in their ads. The label is not a confession. It is a promise printed on the box.
The villain in this story is silence about how the content was made. A coach posts an AI clone reel, says nothing, and it works for a month. Then one follower spots it, posts a screenshot, and the comments fill with one word: fake. Months of trust, gone in a weekend.
What are ASCI’s new AI labelling rules for ads?
On 29 September 2026, ASCI released its Guidelines for Responsible Labelling of Synthetically Generated Content in Advertising. “Synthetically generated content” is the official way of saying content made or changed by AI.
The big idea is simple. ASCI does not care which tool you used. It cares whether AI could change what a buyer understands or decides. In its own words, labelling is mandatory where AI content “materially influences consumer decisions and the absence of disclosure could mislead consumers.”
ASCI also says something every coach should underline: all ads using AI content must still comply with the full ASCI Code. A label is not a free pass.
When do the ASCI AI rules come into effect?
ASCI’s press release says the guidelines take effect three months from the date of publication. Published 29 September 2026, so that is around the end of December 2026.
There is a second set of rules running alongside. Since 20 February 2026, India’s IT Amendment Rules have required platforms to label AI content. MediaNama noted on 6 October 2026 that ASCI and the IT Rules use different tests. So your reel may carry a platform label even where ASCI does not ask you for one. Treat the platform label as a bonus, not as your label.
What must be labelled? Six cases from ASCI
| ASCI says label | What it looks like for a coach or creator |
|---|---|
| AI influencers and ambassadors | A made up face that “recommends” your program |
| A real person’s likeness or voice, with consent | Your AI clone or AI voice selling your course |
| Fabricated events or settings | AI shows a sold out seminar or a studio you don’t have |
| Demos of products that don’t exist yet | An AI preview of an app or planner still being built |
| Key AI sound effects | AI made audio when the sound is the product, like a meditation track |
| Paid AI product suggestions | Must say “Sponsored by [Brand]” |
If you are thinking about a clone of yourself, read what to clone and what not to clone as a coach before you record anything.
What is not allowed, even with a label?
Reality check
“If I add an AI label, I can show anything.”
No. ASCI says ads that are illegal, misleading or break the Code stay prohibited whether or not you label them.
ASCI names five things that stay banned:
- Fabricated endorsements or testimonials
- Exaggerated product results through synthetic images
- Non existent places or settings shown as real to mislead
- Unauthorised copyrighted work
- Deepfakes or anyone’s likeness without consent
For coaches, the top two are the dangerous ones. An AI “student” with a glowing review. An AI before and after. They feel like small shortcuts. They are the fastest way to lose a community. Our guide on fake timers and cure claims covers the same trap from another angle, and here is how to collect real testimonials on WhatsApp instead.
Tool: label it or not?
Twelve real situations coaches and creators ask me about. Tap each card to see the verdict and the reason.
Tool · 12 cards, tap to check
Label it or not?
Make your guess. Then tap. Green means no label, amber means label it, red means not allowed.
1. My AI clone, my face and my voice, with my yes, sells my course in a reel.
LabelA real person’s likeness or voice, replicated with consent, is on ASCI’s must label list. Add “Video created using AI”.
2. AI fixed the lighting and colour in my coaching photo.
No labelRoutine editing like colour correction and light adjustment does not change what a buyer decides.
3. AI wrote the caption for my ad.
No labelASCI lists text uses, such as generating ad copy, as not needing a label. The claims inside still have to be true.
4. An AI made “student” says my course changed her life.
Not allowedFabricated testimonials are prohibited, label or no label. Use real clients’ words, with their permission.
5. AI added Hindi subtitles and translated my caption.
No labelAccessibility uses like subtitles and accurate translations need no label.
6. An AI influencer who does not exist promotes my program.
LabelSynthetic influencers and ambassadors must be labelled.
7. An AI before and after photo shows a body result.
Not allowedExaggerating results through synthetic images is prohibited. A label does not fix a misleading result.
8. Soft AI sparkles and a blurred background behind me.
No labelDecorative or ambient elements do not mislead anyone.
9. AI shows me teaching a packed hall that never happened.
Label, or not allowedA fabricated setting that shapes a buyer’s view needs a label. If it makes people believe the event was real, it is not allowed at all.
10. A deepfake of a famous person praising my course.
Not allowedUsing someone’s likeness without consent is prohibited. Full stop.
11. I paid for an AI chatbot to suggest my course to users.
Label: “Sponsored by”Paid AI product suggestions must say “Sponsored by [Brand]”.
12. A cartoon dragon explains money tips in my reel.
No labelObvious fantasy effects need no label. Nobody thinks the dragon is real.
Verdicts are our reading of ASCI’s categories, for planning. Not legal advice.
The Jaggery Test: three questions before you post
You won’t remember a table at 11 at night when a reel is ready to go. You will remember three questions.
Save this framework
The Jaggery Test: 3 questions before you post
Say what is in the box · sell what is in the box
Save it, or send it to a coach friend
Send on WhatsApp
Label lines to copy
ASCI suggests “Audio/Video created using AI” or “Audio/Video enhanced using AI”, and allows platform labels or any label that accurately tells people what is synthetic. Here are lines that sound human and still say it clearly.
Free kit: 5 label lines
1 · On screen, AI clone reel
Video created using AI. My words, my yes, my AI twin.
2 · Caption line, AI voice
Audio created using AI from my own voice, with my consent. The advice is 100% mine.
3 · Caption line, AI setting
Video enhanced using AI. The background is generated, the coaching is real.
4 · Pinned comment when someone asks “Is this AI?”
Yes, and thank you for asking. I use an AI version of my voice and face so I can answer more of you, more often. Everything it says comes from my own notes. Want the real me? DM me, I reply myself.
5 · Bio line
Some of my videos use an AI version of me, always labelled. The coaching, the care and the replies are mine.
Planning avatar videos for YouTube or Reels? Our guide to AI avatar videos in India covers the setup. And if you run Meta ads, read who is responsible when Meta’s AI edits your ad, because an AI image Meta adds is still your ad under the ASCI Code.
How we help: AI that sounds like you, labelled with pride
Our method, labelled as ours
Founder Voice, the right way
- We start with your consent and your notes. Your voice and face are used where you say yes, for what you say yes to, and nowhere else.
- We build content from your real words. Voice notes and real lessons, never invented results or invented students.
- We label it clearly. On screen and in captions, in lines that sound like you. An AI that feels human and keeps improving itself, and never pretends to be you without saying so.
Thinking about an AI version of you?
Send me a message. Tell me what you make and where you post. I’ll tell you honestly what I would use AI for, and what I would keep human.
WhatsApp Madhur See Founder VoiceWhatsApp opens with a message ready. Nothing is sent until you tap send.
Save this
The AI label card
- Label: AI clone, AI voice, AI influencer, fake setting, future product.
- No label: light and colour fixes, captions, subtitles, obvious fantasy.
- Never: fake testimonials, fake results, deepfakes.
- Say it: “Video created using AI” or “enhanced using AI”.
- Start date: about three months from 29 Sep 2026.
Screenshot this. Check it before every post with AI in it.
Questions coaches and creators ask about AI labels
Do I need to label AI generated content in ads in India?
Under ASCI’s guidelines released on 29 Sep 2026, labelling is mandatory where synthetic content materially influences a buyer’s decision and leaving it unlabelled could mislead. Examples include AI influencers, a real person’s cloned face or voice, fabricated events or settings, and demos of products that do not exist yet.
When do the ASCI AI labelling guidelines come into effect?
ASCI says they take effect three months from the date of publication, which was 29 September 2026. That lands around the end of December 2026.
Can I use an AI avatar of myself in my ads?
Yes, with your consent and a clear label. ASCI lists replicating a real person’s likeness or voice, with consent, as a case that must be labelled. Using anyone’s likeness without consent, including deepfakes, is not allowed.
Are AI made testimonials allowed in ads?
No. ASCI lists fabricating endorsements or testimonials as prohibited, whether or not you add an AI label. Use real client words with their permission.
What label should I use for AI in an ad?
ASCI suggests labels such as “Audio/Video created using AI” or “Audio/Video enhanced using AI”, including labels provided by ad platforms. Any label is fine if it accurately tells people what was synthetic.
Do I need a label for AI written captions or subtitles?
No. ASCI lists text uses such as generating ad copy, and accessibility uses such as subtitles, translations and accurate audio descriptions, as not needing a label.
The sweet shop never hid the jaggery. It wrote it on the box, and the box sold better. Write it on your box.
About this guide: sources, all checked 3 Nov 2026. ASCI press release, “ASCI Releases Guidelines for Responsible Labelling of Synthetically Generated Content in Advertising”, Mumbai, 29 Sep 2026 (what must be labelled, what needs no label, what is prohibited, the suggested labels and the three month start). MediaNama, “India’s AI Ad Labelling Rules: What Brands Must Disclose”, 6 Oct 2026 (how ASCI’s test differs from the IT Amendment Rules 2026). The effective date is our count of three months from 29 Sep 2026. We did not find a separate date notice on ASCI’s site. The 12 example verdicts are our reading of ASCI’s categories, labelled as ours, as are the Jaggery Test and the label lines. This is not legal advice. When in doubt, ask ASCI or a lawyer. Madhur Tyagi is the founder of The Human Algorithm., a growth and automation practice in New Delhi that helps Indian founders, coaches and creators reply faster, follow up kindly and earn trust in their DMs and WhatsApp. 8+ years, 25+ brands, since 2017.
THA Circle
Use AI. Still sound human.
A free WhatsApp channel for founders, coaches and creators who want AI to save time without losing the human touch.
- On joining A welcome pack: 3 AI prompts and 5 chat scripts you can copy today.
- Monday A short script for your own chats. Ten minutes to try.
- Wednesday One real chat taken apart, names removed.
- Friday Human of the Week. A moment a customer felt seen.
Free. Takes ten seconds. Channel followers stay private. Leave any time.
The Sunday note
How people pay attention, and how to earn it. One email a week.
I study AI, attention, human psychology and marketing. Every Sunday I send one idea and one script for your chats. Five minutes to read. Reply any time, I read every answer.
Get the Sunday noteFree. Leave any time. Your email stays private.
Coaches
Found by AI
Coaches