Google is expanding its use of large language models to better detect and block invalid traffic across its ad platforms. It refers to ad interactions that don’t come from real users, wasting ad spend and damaging trust in the ecosystem.
The Ad Traffic Quality team, in collaboration with Google Research and DeepMind, has rolled out new protections that analyze app content, ad placements, and user interactions in real time. This approach has led to a 40% drop in traffic linked to misleading or disruptive ad behavior.
These AI-driven improvements help advertisers avoid wasted impressions and ensure that policy-violating traffic is excluded from billing. Google continues to combine automated and manual systems to maintain accuracy and defend advertisers from evolving click fraud threats.
Read the official announcement here: How we’re using AI in new ways to fight invalid traffic.
