Google's Gemini blocks ads

Google says its Gemini artificial intelligence systems now stop most bad ads before they appear, after blocking or removing more than 8.3 billion ads in 2025. (blog.google.com) The company said on April 16 that Gemini-powered tools caught more than 99% of policy-violating ads before they were shown to users and suspended 24.9 million advertiser accounts last year. (blog.google.com) Google’s 2025 Ads Safety Report says Gemini analyzes “hundreds of billions of signals,” including account age, behavior patterns, and campaign activity, instead of relying mainly on older keyword matching. (blog.google.com) That matters inside Google’s ad business because the company is trying to sort malicious campaigns from legitimate ones with fewer broad account bans. Google said the added precision cut incorrect advertiser suspensions by 80% in 2025. (services.google.com) The report also shows the scale of the cleanup. Google said it removed 602 million scam-related ads, suspended 4 million scam-linked accounts, restricted 4.8 billion ads, and blocked or restricted more than 480 million web pages. (services.google.com) Google said Gemini also helped it process four times as many user reports as in the previous year, giving human reviewers more leads after automated systems had already filtered most violations. (services.google.com) The biggest ad-policy categories by volume included ad network abuse at 1.29 billion ads, personalization violations at 755 million, legal requirements at 646.7 million, and misrepresentation at 421.5 million. (services.google.com) Outside Google’s own framing, the numbers point to a shift in how the company polices ads: more individual ads blocked, fewer blunt account suspensions. TechCrunch reported Google blocked 8.3 billion ads in 2025, up from 5.1 billion a year earlier, while suspensions did not rise at the same pace. (techcrunch.com) Google said advertiser verification remains a second layer of defense alongside Gemini, aimed at stopping bad actors before they buy ads at all. The company is betting that mix of identity checks and machine-led screening can keep harmful ads out without sweeping up paying customers. (blog.google.com)