Google is actively combating invalid traffic (IVT)—non-human generated advertising interactions—through advanced AI and machine learning. Collaborating with Google Research and DeepMind, the company has developed multi-modal language models that mimic genuine user behavior on websites and mobile apps. These tools help identify and flag disruptive ads, hidden methods of ad display, and accidental clicks, leading to a 40% decrease in mobile IVT from deceptive practices in a pilot program from December 2023 to October 2024.
The global IVT rates are significant, with 18% on web and 31% on mobile apps, creating issues for advertisers and publishers alike. Google has already shut down multiple ad fraud schemes affecting millions of users. Its AI-driven approach promises quicker detection and mitigation of fraudulent activities, making the digital ad ecosystem more secure. Continuous improvements in AI efficiency have reportedly led to a 10,000-fold reduction in training data needed to eliminate policy-violating ads.
Source link
