Google blocked 8.3 billion ads in 2025. The company says AI stopped most before users ever saw them.
Big jump in ad blocks, fewer account suspensions
Google said it blocked a record 8.3 billion ads worldwide in 2025, a sharp rise from 5.1 billion the year before. The increase came alongside a drop — not an increase — in the number of advertiser accounts the company suspended. That contrast points to a change in how Google enforces its advertising rules.
In the U.S. The company removed more than 1.7 billion ads and suspended about 3.3 million advertiser accounts last year. In India, which Google identifies as its largest market by users, the firm blocked roughly 483.7 million ads in 2025 while account suspensions fell to 1.7 million from 2.9 million a year earlier. Across all markets, Google tied 602 million of the blocked ads and four million suspended advertiser accounts to scams.
The numbers come from Google’s 2025 Ads Safety Report and public statements from the company. They show a shift away from broad account-level discipline and toward stopping individual policy-violating creatives before they reach people.
AI at the centre of the new approach
Google credited its family of Gemini AI models with making the change possible.
The company said those systems allowed it to detect problematic content at scale and to remove ads earlier in a campaign lifecycle. Google reported that its AI-driven systems stopped more than 99% of policy-violating ads before those ads were shown to users.
Keerat Sharma, vice-president and general manager of ads privacy and safety at Google, told reporters that the company has moved toward more targeted, AI-driven enforcement at a more granular level. The shift, Google says, gives it the ability to block single ads or specific creative elements without necessarily suspending an entire advertiser account.
Google also linked the rise in blocked ads to scammers' use of generative AI. The company said bad actors were increasingly using AI tools to churn out deceptive or misrepresentative ad content at scale. Gemini, Google said, can spot patterns across large campaigns and neutralize them sooner than manual review processes could.
What the numbers mean for advertisers and users
For legitimate advertisers, the change could mean fewer sudden account suspensions and more surgical interventions.
An ad flagged and blocked for a policy violation might no longer trigger a full account suspension if the rest of the advertiser’s activity appears compliant.
That’s likely to be welcome for small businesses that rely on ads and have limited resources for contesting or appealing suspensions. But it could also create new headaches. Advertisers will still need to monitor creative content closely. An individual ad might be blocked for a violation even when the broader campaign remains active.
For users, Google’s approach aims to reduce exposure to scams, misleading claims, and other harmful ad content. The company highlighted ad network abuse, misrepresentation and sexual content among the most common violations it took action on last year.
By intercepting violative ads earlier, Google wants to reduce the likelihood that consumers will see or click on harmful creatives.
How enforcement changed across markets
The pattern of more blocked ads but fewer account suspensions wasn’t uniform across regions, but it was widespread. In India, blocks nearly doubled year over year while suspensions dropped. In the U.S., Google both removed large numbers of ads and carried out millions of account suspensions, showing the company still uses account-level action where it deems necessary.
Google’s report also pointed to common violation categories that vary by country: trademarks, financial services and copyright issues were flagged frequently in India, while ad network abuse and misrepresentation featured heavily in the U.S. The company said these patterns emerged as it combined automated detection with human review.
Automation, scale and the challenge of bad actors
Advertising platforms have long wrestled with the trade-off between scale and control. Platforms need to process trillions of ad impressions and still keep users safe. Automation has been the natural next step — and Google’s latest numbers show the company is doubling down on that strategy.
Generative AI has helped bad actors produce large volumes of believable creative quickly. Google described the problem as one of scale: when scammers can generate thousands of slightly different ads in minutes, human review can’t keep up. The company argues AI can detect subtle patterns across those variants and cut campaigns off earlier.
That capability matters because stopping a single creative is less disruptive to an advertiser than suspending the whole account. It also reduces collateral damage, at least in theory, by avoiding penalties for advertisers whose violations are limited to a handful of ads.
Trade-offs and unanswered questions
The new approach raises important questions about transparency and redress.
A more granular, automated system can act faster. It can also be harder for advertisers to understand why an individual ad was blocked if the decision rests on pattern recognition across many assets.
Google’s report and its public remarks didn't provide exhaustive detail about appeal timelines, the percentage of automated decisions that get reversed on review, or how advertisers are notified when the company blocks specific creatives. Those operational details matter to brands and agencies that run large, complex campaigns.
Regulators are also paying closer attention to how platforms use AI for content moderation and enforcement. Governments around the world have been asking platforms to explain enforcement processes and provide clear paths for appeal. Google’s pivot toward ad-level blocking is likely to attract scrutiny in jurisdictions focused on platform accountability.
What comes next for ad safety
Google said it will keep integrating Gemini and other AI tools deeper into its advertising systems, both to create and to police ads. The company positions that work as a dual effort: using AI to automate legitimate ad creation while using the same class of models to detect and stop abuse.
That dual use raises a simple question for the industry: can the same technology that helps marketers scale also keep bad actors in check? Google’s 2025 figures suggest the company believes the answer is yes — at least so far. How well that holds up will depend on whether AI detection can stay a step ahead of the new, automated tactics bad actors adopt.
Advertisers will need to adapt. They’ll have to ensure creative workflows, approval processes and compliance checks catch violations before launch.
Agencies will have to explain to clients how ad-level blocks differ from account suspensions, and how to contest them. Platforms will have to make those processes clearer.
Related Articles
Google said its AI-driven systems caught more than 99% of policy-violating ads in 2025 before they were shown to users.
This article was created with AI assistance.
