AI news · ADEX · Ad fraud
ADEX says AI played a part in 52% of ad fraud schemes
The anti-fraud firm says cloned brands and fake endorsements turned up in over half of the AI-assisted schemes it found on eight ad platforms this year.
Of the ad fraud schemes ADEX flagged between January and September, 52 per cent made significant use of AI, the anti-fraud business of AdTech Holding said on 27 October 2026 (ADEX’s release on Hackread). Its State of Ad Fraud 2026 report rests on moderation records from eight ad platforms that run ADEX, spanning 4.9 million creatives in 1.2 million campaigns.
Within the AI-assisted schemes, bots posing as people showed up in 64 per cent, deepfakes in 11 per cent, and cloned brands or fake endorsements in 53 per cent; the counts overlap, so they add up to more than 100. ADEX expects fraudsters to aim at AI shopping agents in 2027, with ads written to mislead the software that browses and buys for people.
PPC Land points out that the 52 per cent is measured per scheme, and that each scheme can cover many campaigns. The release leaves the eight platforms unnamed and gives no total of schemes, and PPC Land adds that the figures come from detection products ADEX itself sells.
Our rights register, read on 24 September 2026, covers 12 AI video generators. On their consumer plans every one makes showing another real person conditional on that person’s consent or clearance, Seedance at least outside the US, where its US terms are unclear; Google’s Veo terms add a ban on impersonation meant to deceive.
So a fake celebrity endorsement made with any of those tools already breaks the generator’s own terms, and the check that catches it, in ADEX’s figures, is the ad platforms’ moderation.
Sources
- PPC Land: AI was involved in 52% of ad fraud schemes ADEX flagged in 2026. Checked
- Hackread: ADEX: AI Involved in 52% of Detected Ad Fraud Schemes. Checked
Alexandre S.
Alexandre S. is a creative ad strategist and copywriter with ten years' experience in digital marketing across multiple channels.