Google reports huge ad‑safety takedowns

Google says its ad-safety systems blocked or removed more than 8.3 billion ads worldwide in 2025, with more than 99% stopped before anyone saw them. (blog.google) The company released the figures on April 16, 2026, in its 2025 Ads Safety Report. Google also said it suspended 24.9 million advertiser accounts, restricted 4.8 billion ads, blocked or restricted 480 million web pages, and took action on 245,000 publisher sites. (blog.google) (services.google.com) In India, Google said it blocked 483.7 million ads and suspended 1.7 million advertiser accounts in 2025. The top violation categories there were trademark, financial services, copyright, personalization, and abuse of the ad network. (moneycontrol.com) Google says Gemini, its family of artificial intelligence models, now reviews ads by reading patterns instead of just matching keywords. The company said those systems look at signals such as account age, behavior, and campaign patterns to catch scams earlier. (blog.google) That shift changed what Google enforced. TechCrunch reported that blocked ads jumped to 8.3 billion in 2025 from 5.1 billion in 2024, while advertiser suspensions fell from 39.2 million to 24.9 million worldwide. (techcrunch.com) (blog.google) Google told reporters it is acting “at a much more granular level” by blocking specific ads instead of relying as often on account-wide bans. Keerat Sharma, Google’s vice president and general manager for Ads Privacy and Safety, said that approach cut incorrect advertiser suspensions by 80% year over year. (techcrunch.com) (services.google.com) Scams remained a major category inside the broader crackdown. Google said 602 million ads and 4 million advertiser accounts were tied to scams in 2025. (blog.google) (services.google.com) Google also said it made 35 ad-policy updates in 2025. The report lists some of the biggest enforcement buckets as abuse of the ad network, personalization violations, legal requirements, misrepresentation, trademark, and financial services. (services.google.com) The company says generative artificial intelligence is now helping both sides of the fight. Google said scammers are using AI to produce deceptive ads at scale, while its own systems are reviewing the majority of Responsive Search Ads instantly and blocking harmful submissions before they run. (blog.google) The report leaves Google with a narrower target than the headline numbers suggest: catch the remaining ads that slip through without sweeping up legitimate advertisers by mistake. That is now the company’s stated test for whether its bigger takedown totals actually translate into a cleaner ad system. (services.google.com)