Deepfake Detection Tool Catches Fake YouTube Ads Using Real Faces

YouTuber Michel Janse found her face in an advert she never agreed to. Someone took her videos, built a deepfake, and ran it as an ad. She did not shoot it. She did not approve it. She found out like everyone else: by seeing it.

How it worked

Face-swap deepfakes need source footage, and YouTubers publish hours of it. The scammers lifted her face from her own videos, mapped it onto an actor, and rendered a convincing ad. Then they bought ad placements and ran it.

This is the creator economy's new nightmare. Your face is your business, and anyone with a GPU can now wear it. The ads often push shady products, which means the victim's reputation takes the hit twice: once for the theft, once for the junk being sold.

Detection angle

A deepfake detection tool can flag these ads by analysing facial inconsistencies: blinking patterns, edge artefacts around the jawline, mismatched lighting between face and background. Fake video detection online is getting fast enough to scan ad creatives at scale.

If you are a creator, search for your own face regularly. If you run an ad platform, a deepfake video detector online should be part of your creative review. The platforms profit from every ad served, real or fake. That has to change. Nobody should find out they starred in an ad from the comments section. Most victims never find out at all. Janse at least got to fight back.

Your face, your rules. Try the AI Detector on any ad that looks off.