Google has released its 2025 Ads Safety Report, highlighting large-scale enforcement actions across its advertising ecosystem, driven by AI-led detection systems.
Globally, the company has blocked or removed over 8.3 billion ads, with more than 99% stopped before being served to users. It has also suspended 24.9 million advertiser accounts and restricted 4.8 billion ads during the year.
Enforcement has extended beyond ads, with action taken on over 480 million web pages, and 245,000 publisher sites having faced policy enforcement measures. The company has also introduced 35 policy updates in 2025.
AI systems have played a central role in enforcement. Out of the total pages actioned, over 467 million pages have been detected and enforced through AI-driven systems, contributing to a detection rate exceeding 97% in key violation categories such as sexual content.
Category-wise enforcement has shown significant volumes. Ads related to abusing the ad network (1.29 billion), personalisation violations (755 million), legal requirements (646.7 million), and misrepresentation (421.5 million) have accounted for a large share of removals. Other key categories have included trademark violations, financial services, sexual content, gambling, and healthcare-related ads.
The report has also outlined action against scams, where over 602 million ads have been removed for violations associated with fraudulent activity, and nearly 4 million advertiser accounts have been suspended.
In India, Google has removed 483.7 million ads and suspended 1.7 million advertiser accounts in 2025. The top policy violations in the country have included trademark, financial services, copyright, personalisation violations, and abuse of the ad network.
The report has indicated that improvements in AI systems have enabled more precise enforcement, resulting in an 80% reduction in incorrect advertiser suspensions, while maintaining high detection rates.
The company’s advertiser verification program has continued to act as an additional layer of control by validating advertiser identities and limiting misuse of the platform.
