Google: 8.3B harmful ads blocked

Google said it blocked or removed more than 8.3 billion ads in 2025, with Gemini-powered systems stopping most policy violations before those ads ran. (blog.google) The company said it suspended 24.9 million advertiser accounts last year, including 4 million accounts tied to scams, and blocked or removed 602 million scam-related ads. Google said its systems now stop more than 99% of policy-violating ads before users see them. (blog.google) Google published the figures in its 2025 Ads Safety Report on April 16, 2026, six days after the post appeared on The Keyword. The report frames generative artificial intelligence as a screening tool that helps detect fraud patterns, fake business signals and repeat bad actors at scale. (blog.google) Online ad safety is the business of checking ads and advertiser accounts before they reach search results, websites or YouTube. Google’s latest report shows the company is putting more of that review work into automated systems built on Gemini, its family of artificial intelligence models. (blog.google) Google paired the safety report with a broader push around Ads Advisor, a Gemini-based assistant inside Google Ads. The product is in beta and can answer questions, troubleshoot performance and policy issues, and suggest new text and image creatives. (support.google.com) Google has also been pitching Ads Advisor as a tool that acts on an advertiser’s goals, not just a chatbot that explains settings. In November 2025, Google said Ads Advisor and Analytics Advisor were rolling out to all English-language accounts and would be available globally in early December. (blog.google) The safety numbers also extend a recent trend in Google’s annual reports: more emphasis on prevention than cleanup after the fact. In its 2024 Ads Safety Report, published April 16, 2025, Google said artificial intelligence was improving its ability to stop fraudsters before they could show ads to people. (blog.google) That matters for advertisers as well as users, because policy enforcement can freeze campaigns, suspend accounts and delay launches when systems misread intent. Google’s help pages describe Ads Advisor as a way to surface policy-related issues and guide advertisers through fixes inside the ad platform. (support.google.com) Google’s pitch is that the same Gemini systems helping marketers write ads and analyze campaigns can also police abuse and protect accounts. The company is now selling artificial intelligence in ads as both a growth tool and a guardrail. (blog.google)