Google blocked a record 8.3 billion ads globally in 2025, up from 5.1 billion the previous year. But in a surprising twist, the company suspended far fewer advertiser accounts than that surge might suggest. The reason: Google is fundamentally changing how it polices its advertising platform, shifting from banning bad actors outright to blocking individual ads with surgical precision using its AI models.
The Numbers Behind the Shift
The figures come from Google's 2025 Ads Safety Report, released Thursday. The company said its AI-driven systems powered by its Gemini models caught more than 99 percent of policy-violating ads before they were ever shown to users. That is an extraordinary detection rate, and it explains why the volume of blocked ads surged so dramatically year over year.
Among the blocked ads and suspended accounts, 602 million ads and 4 million advertiser accounts were linked to scams. In the United States alone, Google removed over 1.7 billion ads and suspended 3.3 million advertiser accounts, with ad network abuse, misrepresentation, and sexual content among the most common violations.
India, Google's largest market by users, saw 483.7 million ads blocked nearly double the previous year even as account suspensions actually fell from 2.9 million to 1.7 million. Trademarks, financial services, and copyright issues were the top violations in the Indian market.
Why Fewer Bans Despite More Blocks
Keerat Sharma, VP and general manager of ads privacy and safety at Google, explained the logic at a virtual briefing. The company has shifted toward more targeted, AI-driven enforcement at a granular level, blocking individual ads rather than using blunt instruments like full account suspensions. This approach has helped reduce incorrect suspensions by 80 percent year over year.
The strategy reflects a fundamental rethinking of enforcement. Instead of punishing entire accounts which can inadvertently catch legitimate advertisers who made minor policy mistakes Google is now using AI to identify and remove specific problematic ads while leaving the rest of the advertiser's campaigns intact.
Google's layered defenses also include advertiser verification, a process that requires businesses to confirm their identity before running ads. Sharma said this has helped prevent bad actors from creating accounts in the first place, which contributes to the decline in overall suspensions.
Generative AI Is a Double-Edged Sword
The report also highlights a growing challenge: scammers are increasingly using generative AI to produce deceptive content at scale. Fraudulent ads that once required manual effort to create can now be generated in seconds using widely available AI tools, making the volume of bad ads much harder to manage.
Google says its Gemini models are key to countering this trend. The AI systems can detect patterns across large campaigns and block them earlier in the pipeline before the ads reach users. It is essentially an AI-versus-AI arms race, with Google deploying its own models to identify and stop the output of models used by bad actors.
This dynamic mirrors broader cybersecurity concerns across the tech industry, where AI is simultaneously making attacks easier to launch and defenses more powerful to deploy.
The India Factor
The Indian market deserves special attention. With nearly half a billion ads blocked in a single year, India has become one of Google's most challenging enforcement environments. The sharp increase in blocked ads, combined with the decline in account suspensions, suggests that Google is applying its new AI-first enforcement model aggressively in the market.
Given that Google recently expanded its Gemini features in India and launched personal intelligence capabilities for Indian users, the country is clearly a priority for both product expansion and platform safety.
What Comes Next
Sharma acknowledged that the numbers will likely fluctuate as Google rolls out new defenses and bad actors adapt their strategies. The company's goal is to stop harmful ads as early in the pipeline as possible ideally before they are even submitted for review.
The broader trend is clear: AI is reshaping how the world's largest advertising platform enforces its own rules. Whether this precision-based approach proves more effective than traditional account-level enforcement will become clearer in the years ahead. For now, 8.3 billion blocked ads in a single year is a number that speaks for itself.
