Google released its 2025 Ads Safety Report on Thursday with a notable finding: the company blocked 8.3 billion ads globally last year — a 63% increase from 5.1 billion in 2024 — yet suspended far fewer advertiser accounts than that surge might suggest.
The disparity reflects a deliberate strategic shift in how Google polices its platform. Rather than banning problematic advertisers outright, the company is increasingly using AI to block individual violating ads before they reach users. Gemini models, Google's family of AI systems, caught over 99% of policy-violating ads last year before display.
Google attributes the rise partly to how scammers now operate: generative AI allows bad actors to produce deceptive content at scale. Gemini's pattern-detection capabilities help Google identify and block these campaigns earlier in the pipeline.
The numbers are instructive. Among the 8.3 billion blocked ads, 602 million and 4 million advertiser account suspensions were linked to scams. In the United States, Google removed over 1.7 billion ads and suspended 3.3 million advertiser accounts, with ad network abuse, misrepresentation, and sexual content as top violations. In India — Google's largest market by users — the company blocked 483.7 million ads, nearly double the prior year, while account suspensions fell to 1.7 million from 2.9 million.
Keerat Sharma, VP and general manager of ads privacy and safety at Google, framed the shift explicitly during a virtual briefing: the company has moved toward "more targeted, AI-driven enforcement at a much more granular level, on a creative level, as opposed to using a much more blunt instrument, like advertiser suspensions." That precision has tangible results — incorrect suspensions dropped 80% year over year.
Google's layered approach includes advertiser verification, which requires businesses to confirm identity before running ads. Sharma said this mechanism helps prevent bad actors from creating accounts in the first place, contributing to the decline in suspensions. The company expects the numbers to fluctuate as it rolls out new defenses and bad actors adapt, with the goal of stopping harmful ads as early as possible in the pipeline.