Google now targets bad ads, not bad actors

Google announced Thursday that it will block a record 8.3 billion ads worldwide in 2025. This was an increase from 5.1 billion the previous year. But the number of advertiser accounts the company has suspended is far smaller than this spike suggests, raising questions about how it manages the platform.

The search giant blames this disparity on its growing use of AI, particularly its Gemini model (Google’s family of AI systems). Google says this will allow it to detect and block ads that violate its policies earlier and with greater accuracy. Last year, the company said its AI-driven systems caught more than 99% of these ads before they were shown to users.

Both findings come from Google’s 2025 Ad Safety Report and reflect broader changes in enforcement. While more problematic ads are being suspended, the number of suspended advertiser accounts is decreasing, suggesting a focus on blocking individual ads.

Google said the increase in blocked ads also reflects the increasing use of generative AI to help fraudsters create deceptive content at scale, and said its Gemini model helps detect patterns across large campaigns and block them early.

The move reflects a broader push by Google to integrate its Gemini model more deeply into its core products and infrastructure, such as advertising, as the company increases its use of AI to automate campaign creation, detect policy violations, and respond to emerging threats in real-time.

Of the ads blocked and accounts suspended, 602 million ads and 4 million advertiser accounts were related to fraud, the company said.

Google removed more than 1.7 billion ads and suspended 3.3 million advertiser accounts in the U.S. in 2025, with the most common violations involving ad network abuse, misrepresentation, and sexual content. In India, Google’s largest market with a large number of users, the company blocked 483.7 million ads (nearly double the previous year), even as account suspensions fell from 2.9 million to 1.7 million. Top violations included trademark, financial services, and copyright issues.

Keerat Sharma, Google’s vice president and general manager of advertising privacy and safety, told reporters in a virtual briefing that the company is moving toward more targeted, AI-driven enforcement “at the creative level, at a more granular level, rather than using more blatant measures like suspending advertisers.” He added that this approach helped reduce unwarranted suspensions by 80% year-over-year.

Sharma added that Google’s defenses-in-depth measures, including advertiser verification (a process in which businesses ask for identity verification before running an ad), are designed to prevent bad actors from creating accounts in the first place, which is contributing to the reduction in suspensions.

Sharma said Google aims to stop harmful ads as early in the pipeline as possible, and this number will likely change over time as Google rolls out new defenses and bad actors adapt.