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

Google is expanding its use of Gemini AI models to detect and block malicious ads on its advertising platform. In 2025, the company blocked 8.3 billion ads, suspended 24.9 million advertiser accounts, and removed 602 million scam-related ads. Unlike previous keyword-based detection, Gemini analyzes billions of signals including advertiser behavior, account history, and campaign intent to identify harmful ads in real time. The majority of Responsive Search Ads are now reviewed instantly at submission. Google also reports that improved AI accuracy has reduced incorrect advertiser suspensions by 80%. The move comes as cybercriminals increasingly use generative AI to build more sophisticated malvertising campaigns at scale.