How do I stop fake AI endorsement ads using my face?
Key takeaways
- An ad creative is reviewed separately from an organic post, so reporting only the post leaves the ad running and the spend continuing.
- India's amended IT Rules require platforms to remove AI-generated content within three hours of notification, which is the rule to cite in a written notice.
- Meta's Ads Library and Google's Transparency Report show paid ads by advertiser, and both are searchable without an account in most cases.
- These campaigns are usually delivered through bought, rented or hijacked accounts, so expect the ad to reappear under a new advertiser within days.
- A visible AI-labelling disclosure on your own content helps, because the amended rules make AI labelling mandatory for synthetic content.
As of 24 October 2026: the service descriptions, prices and turnaround figures on this page are current.
Why do fake AI endorsement ads keep coming back?
Because the operation is a business, not a one-off. The creator sells a bundle of a face, a voice and a script, the script is generated per audience, and the ad is delivered through bought or compromised accounts so that when one advertiser is banned, another takes over. Because it is scripted, the same face appears in several unrelated product categories within weeks. That is also the evidence to use: a cluster of ads claiming to be you for products you have never touched is not a misunderstanding, it is a commercial impersonation operation, and documenting the cluster strengthens every individual takedown complaint and any trademark or copyright claim you make afterwards.
Which platforms and routes should I use?
Three routes, in this order. First, the ads route: report the ad in the platform's ads manager, which opens a creative-level review that is separate from the post review and is where the budget and the delivery are actually stopped. Second, the content route: report the post, page or account, and ask for the AI-generated label to be applied. Third, the legal route: a trademark complaint for your name or brand, and a copyright or personality-rights complaint for the likeness itself, which is available to Indian creators and businesses. For Meta, a formal appeal against a declined takedown escalates to a human reviewer. Use a written notice that quotes the three-hour AI-content removal rule in the amended IT Rules.
What evidence should I collect before reporting?
Collect eight things for every single ad. A full-page screenshot showing the URL, the advertiser name, the profile image, the date and time, and the system clock. A cropped screenshot of the ad creative itself. The advertised landing page URL, and a screenshot of the page. The ad library entry, if you can find the same ad in the Ads Library with the advertiser's name. The claim that makes it actionable: that you have never endorsed that product, or the date range during which you did not. And a log of what you reported, when, to whom, and with what reference number. Store everything off-platform in two places, and keep the log because you will need it to show a pattern rather than a one-off.
What can I do to reduce the damage in the meantime?
Four moves. Post a short, clear statement on your own channels stating that you do not endorse certain products and that you will never ask for money or OTP. Report the impersonating accounts for blue-tick or verification, because verification requests are the usual first step in these operations. Publish your own AI-labelling disclosure, which the amended IT Rules make mandatory for synthetic content, so that your genuine content is distinguishable. And monitor the ads libraries daily for two weeks, because an unattended campaign can spend a significant amount before the first complaint lands. If a scam has collected money from your followers, treat that as a fraud case and report it on cybercrime.gov.in and helpline 1930.
| Surface | What the scam looks like | What actually stops it |
|---|---|---|
| Meta ad library | A fake endorsement running as a paid ad | A trademark and likeness notice to Meta, plus a report from the ad library |
| Search and display ads | A fake brand endorsement on a site you visit | The same notice, plus a trademark claim with the platform's brand policy team |
| Marketplaces | A product listing using your likeness | A rights complaint to the marketplace, and a takedown to the seller |
| Deepfake video on YouTube | A cloned voice and face endorsing a product | A copyright and likeness complaint, plus a platform policy report |
| Your own channels | Someone posting as you | A fast password reset, a hardware second factor, and a session revocation |
What are the frequently asked questions?
Get a straight answer on this
Send us the ad URLs you have collected and we will return a fixed-price takedown and ads-monitoring scope with a start time.
Send us the details See pricing