The ad fraud landscape changed sharply in 2026. That is the central finding of the State of Ad Fraud 2026: ADEX Trust & Safety Report, published by anti-fraud platform ADEX.
The report analyzes January–September 2026 moderation and verification data from eight ad platforms, covering 1.2 million campaigns and 4.9 million creatives. Below are the key findings.
Key Findings at a Glance
- 52% of distinct fraud schemes showed AI-generated or AI-assisted assets.
- 64% of AI-involved schemes used bots that pass as human visitors.
- 53% of AI-involved schemes used cloned brands and fake endorsements.
- 11% of AI-involved schemes used deepfakes.
- About one-third of rejections were linked to infected domains, which became the top rejection reason from Q2 onward.
- 6% of Q3 rejections were prize and bonus scareware (“You won a smartphone”), which emerged as its own category that quarter.
What Moved in 2026
According to ADEX, the clearest year-on-year shift is in why platforms suspend advertisers.
| Metric | 2025 / Q1 2026 | Q3 2026 |
|---|---|---|
| Cloaking (share of suspensions) | ~75% | ~45% |
| Malicious iOS scripts (share of suspensions) | Not tracked | ~20% |
| Copyright (share of rejections) | 8% | 21% |
| Region-restricted content (share of rejections) | 5% | 11% |
| Adult content (share of rejections) | 42% | 14% |
Cloaking remains the top suspension reason, but it now shares the list with two newer threats: malicious iOS scripts and messenger-based phishing on services such as WhatsApp and Telegram.
In those messenger schemes, the persuasion happens in the chat itself. ADEX notes that language models let a single operator run hundreds of believable conversations in any language.
Quality Over Quantity
ADEX reports that the number of rejected or blocked campaigns fell by roughly a third after Q1. The report says this does not signal a quieter market. Operators shifted from mass submissions to fewer, better-built campaigns.
“One well-made scam with a cloned brand page does more damage than a hundred crude ones,” said Sergey Martianov, CPO at ADEX.
High-Risk Regions
The ADEX report identifies the regions with the highest share of campaigns rejected at moderation in each period:
- Tier-1 markets (US, UK, Canada, Australia, Germany): high-risk in every quarter
- South Asia: from Q2 onward
- Turkey: first half of the year
- Latin America: Q1
- East Asia: Q3
Industry Context
ADEX places its findings alongside wider industry data cited in the report:
- Thales 2026 Bad Bot Report: bots made up over 53% of internet traffic.
- HUMAN: monthly AI-driven traffic grew 187% during 2025.
- Google’s 2025 Ads Safety Report: more than 8.3 billion ads were blocked or removed.
2027 Outlook
ADEX lists five predictions:
- Ads written for AI agents, using prompt injection
- Agent traffic that blurs the line between bot and buyer
- Fraud operations generated end to end by AI
- More borrowed trust through hacked domains and messenger funnels
- Scareware moving to connected TV, in-app and AI assistant interfaces
ADEX labels prompt injection a forecast, not a category currently measured in its moderation data.
The complete dataset and methodology are available in the full ADEX report.
