The ADEX Trust & Safety team has published its annual report on ad fraud and policy abuse. It covers January to September 2026 and is based on moderation and verification data from eight ad platforms that use ADEX: 1.2 million campaigns and 4.9 million creatives, compared with 2025, where comparable data is available.
The headline finding: fewer policy-violating campaigns reached platform moderation this year, but the ones that did were more complex. Operators moved away from high-volume submissions toward fewer, better-disguised campaigns, and generative AI became part of the standard toolkit.
Key findings
- Cloaking is losing ground. It remains the top reason platforms suspend advertisers, but its share fell from about 75% of suspensions across 2025 to under half in Q3 2026.
- iOS became a target in its own right. Malicious iOS scripts appeared as a separate suspension category in Q3 and accounted for about one in five suspensions that quarter. For background, see our breakdown of the Coruna iOS exploit kit.
- Copyright violations are rising. Their share of campaigns rejected at platform moderation grew from 8% in Q1 to 21% in Q3.
- Adult content is declining. It fell from 42% of rejections in Q1 to 14% in Q3, the largest drop among rejection reasons.
- AI is now common. About half of the abuse schemes ADEX observed (52%) showed indicators of AI-generated or AI-assisted assets.
- Phishing moved into messengers. Messenger-based phishing became a separate reason for platform suspensions, with the persuasion happening in the chat.
Fewer rejected campaigns do not mean a quieter market: bad actors are submitting fewer, better-built campaigns. One well-made scam with a cloned brand page does more damage than a hundred crude ones.
What the report covers
- A step-by-step breakdown of one iOS exploit scheme that ran through about 50 advertiser accounts and targeted Blockchain wallets
- Where AI shows up in abuse: bots that pass as people, cloned brands, deepfake endorsements, and chat funnels
- High-Risk GEOs by quarter, from Tier-1 markets to South Asia, Latin America, and East Asia
- The 2027 outlook: abuse aimed at the AI agents that browse and click on behalf of people, and how to tell good agent traffic from bad (related reading: how an AI agent can turn into an attacker’s delivery tool)
- Five practical checks advertisers and publishers can run now
Download the full report
The full report is available as a PDF. You can read it below or download it.