Google’s AI blocked over 39 million suspicious ad accounts in 2024
In 2024, Google significantly ramped up its ad fraud prevention efforts, using AI to block over 39 million suspicious accounts before their ads could go live.
In 2024, Google blocked 39.2 million accounts on its advertising platform — more than triple the number from the previous year.
The increase is credited to a major anti-fraud initiative, with large language models playing a central role.
AI helped detect fake accounts by identifying patterns such as fraudulent payment information and impersonation attempts. As a result, Google claims it can now block most malicious actors before their ads even go live.
The company upgraded its AI systems more than 50 times and assembled a cross-functional team of over 100 experts — from Ads Safety to DeepMind. Together, they tackled fake ad schemes, including celebrity scam images, and implemented more than 30 new policies for advertisers and publishers.
The results were swift: over 700,000 accounts were banned for violating policies, and user complaints about fake ads dropped by 90%.
