Google expands Gemini AI use to fight malicious ads on its platform

Google reports that it is using its Gemini AI models to detect and block malicious ads at scale, having removed or blocked 8.3 billion ads and suspended 24.9 million advertiser accounts in 2025 (including 602 million ads tied to scams). The report describes malvertising tactics such as cloaking, URL redirects, fake login pages, trojanized software distribution, and crypto-draining sites, and notes attackers are increasingly leveraging generative AI to scale deceptive campaigns; Google says the AI has improved real-time detection and reduced false suspensions.