Google has been deploying its Gemini AI models quietly and effectively to combat ad fraud and invalid traffic (IVT) across web and mobile platforms, leading to notable improvements since late 2023. The company’s ad traffic quality division collaborates with Google Research and DeepMind experts, utilizing multi-modal large language models designed to behave like human users when navigating websites and apps. This approach allows the AI to interact with site elements, observe ads, take screenshots, and detect when ads are deceptively hidden or disruptive.
The Gemini-powered system flags problematic ad behaviors such as accidental clicks, hidden ads that are invisible to users but still register impressions, and intrusive ads that force unwanted interaction. These detections then undergo human review to ensure enforcement accuracy. A pilot program running from December 2023 to October 2024 resulted in a 40% decrease in mobile invalid traffic caused by deceptive or disruptive ads, accelerating enforcement speed and precision.
Marketing Technology Insights: Pollo AI Launched Its Next-Gen Photo to Avatar Video Generator
Invalid traffic remains a severe challenge for the digital ad ecosystem. For context, in Q1 2025, global IVT rates were around 18% on the web and as high as 31% on mobile apps. This traffic not only wastes advertisers’ budgets but also siphons revenue from legitimate publishers and degrades user experience, sometimes even disrupting critical actions like emergency calls or introducing malware risks.
Google’s proactive efforts this year have included removing hundreds of apps linked to large-scale botnet ad fraud operations. It is also pursuing legal actions against fraud rings infecting millions of Android devices. Despite the ongoing cat-and-mouse nature of ad fraud, Google believes AI tools in particular Gemini’s ability to quickly adapt offer a more robust defense compared to traditional data analysis alone.
Marketing Technology Insights: SEO for Lawyers, LLC Expands To Help Attorneys Dominate Local Search
In addition to ad fraud detection, Google is applying Gemini AI in other areas like scam prevention in Search and spam detection on Maps. The company recently reported a dramatic reduction in the amount of AI training data required to detect policy-violating ads, demonstrating gains in efficiency and effectiveness.
Overall, Google’s deployment of Gemini AI represents a significant technological commitment to preserving ad ecosystem integrity, protecting advertisers, publishers, and users alike from the costly and harmful impacts of invalid traffic and online scams. This work is ongoing as both fraudsters and AI defenses evolve rapidly.
Marketing Technology Insights: Online Advantages Launches AI-Driven Search Visibility Services for Local Businesses
For media inquiries, you can write to our MarTech Newsroom at sudipto@intentamplify.com
