Ad fraud strategies have taken a sharp turn in 2026. A new report from the anti-fraud firm ADEX analyzing data from 1.2 million ad campaigns across eight platforms shows that over half of fraud schemes are now using AI-generated or AI-assisted content. At the same time, traditional tactics like cloaking are declining, while threats like iOS-based exploits and messenger phishing are emerging as major risks.
AI Tools Dominate Fraud Trends
Between January and September this year, 52% of detected fraud schemes involved AI-created or AI-enhanced assets. Of those, nearly two-thirds used bots masquerading as real users. A little over half employed cloned brand materials and fake endorsements to appear legitimate, and around one in nine schemes used deepfake technologies.
The New Threat Landscape: iOS Exploits & Messenger Phishing
While cloaking—where ads are hidden from reviewers—used to account for roughly three-quarters of advertiser suspensions, its share has dropped to about 45% by the third quarter. New threats are filling the gaps. Malicious scripts targeting iOS now make up around one-fifth of suspensions, and phishing campaigns delivered via messaging apps like WhatsApp and Telegram are on the rise. These messenger-based frauds often use AI to generate persuasive conversations in many languages.
Higher Quality, Harder-to-Catch Schemes
Though campaigns rejected or blocked fell by nearly a third post-Q1, that’s not cause for optimism in ad safety; fraudsters appear to be shifting toward fewer but more polished attacks. Brand cloning, stronger visual assets, and better script execution are replacing mass submissions of low-quality ads.
Where Ad Fraud Is Most Intense
The report highlights which regions see the most content moderation rejections. Tier-1 markets—including the US, UK, Canada, Australia, and Germany—consistently lead in risk. South Asia’s fraud activity has surged since mid-year, while Turkey, Latin America, and East Asia also show elevated levels at different points.
Looking Ahead to 2027
ADEX predicts several emerging trends for the rest of this year and beyond:
- Ads crafted for AI agents, including those vulnerable to prompt injection.
- Agent-driven traffic that blurs the distinction between bots and genuine buyers.
- Full‐scale fraud operations designed and run by AI from start to finish.
- More use of hacked domains and messenger funnels to borrow trust.
- Scareware expanding into connected TV, in-app experiences, and interactions with AI assistants.
- Note: prompt injection is viewed as a forecast rather than a currently measured threat category.
Broader industry figures back up these changes. Reports indicate bot traffic now accounts for over 53% of all internet traffic, AI-driven traffic spiked by approximately 187% in 2025, and billions of ads are being removed for safety violations.
Why it matters: The evolution in ad fraud shows that fraudsters aren’t just increasing volume—they’re upping their sophistication. As AI becomes more accessible, forged identities, cloned brands, and deepfakes pose growing risks to both businesses and consumers. Ad platforms, marketers, and regulators must adapt quickly: enhancing detection tools, pushing for better verification, and preparing for new battlegrounds such as messenger apps and AI agents. This shift doesn’t just change tactics—it forces a rethink of how trust and authenticity are built online.
