Google on Thursday said its safety teams work around the clock to stop bad actors using increasingly sophisticated malicious ads. In 2025, Gemini-powered tools improved detection and blocking of harmful ads before they were served.
According to Keerat Sharma, VP & General Manager, Ads Privacy and Safety, the systems now catch over 99% of policy-violating ads before they are shown, as Google continues evolving its defenses to counter advanced scam tactics.
Gemini Improves Ad Safety and Detection
Google uses AI systems to detect and block scammers, and Gemini expands this capability through deeper analysis of advertising behavior. The models process hundreds of billions of signals, including account age, behavioral cues, and campaign patterns, to identify suspicious activity before ads reach users.
Unlike earlier keyword-based systems, Gemini focuses on intent understanding, allowing detection of malicious ads even when they are designed to bypass traditional filters. This enables earlier and more accurate blocking within the ad delivery process.
Core Updates from Google’s Ad Safety Report
Google reported large-scale enforcement actions in 2025 as Gemini improved detection speed and accuracy across its ad systems. The company blocked or removed billions of harmful ads and strengthened automated review systems along with user report handling.
Advertiser verification also remains an important preventive layer, helping block bad actors before they enter the system and improving overall ad ecosystem integrity.
- Over 99% of policy-violating ads blocked before serving in 2025
- More than 8.3 billion ads blocked or removed
- Around 24.9 million advertiser accounts suspended
- Including 602 million scam-related ads and 4 million scam-linked accounts
- Gemini enables real-time detection of AI-generated deceptive ads
- Most Responsive Search Ads reviewed instantly, with harmful content blocked at submission
- Expansion of instant review planned for more ad formats
- User report processing increased by 4x in 2025 vs previous year
- Incorrect advertiser suspensions reduced by 80% last year
- Advertiser verification adds identity checks before ads enter the system
Outlook
Google says Gemini will continue strengthening its ability to detect and block evolving scam techniques, especially those powered by generative AI. The company is expanding intent-based detection, real-time enforcement, and verification systems to improve prevention and response while maintaining support for legitimate advertisers.