Necessary Always Active
Necessary cookies are required to enable the basic features of this site, such as providing secure log-in or adjusting your consent preferences. These cookies do not store any personally identifiable data.
|
||||||
|
||||||
|
||||||
|
Published on
April 16, 2025
5 min read
Google Uses AI to Block 39 Million Fraudulent Ad Accounts in the U.S.
Google blocked 39.2 million advertiser accounts in 2024 as part of its ongoing fight against ad fraud, according to a TechCrunch report published on Wednesday. This figure represents more than a threefold increase compared to the previous year’s data. Google’s use of Large Language Models (LLMs) helped detect fraudulent advertising at its source, thereby eliminating misleading or harmful ads.
Advanced AI tools identified both fake businesses and abnormal payment activity to block the accounts. Google used AI to suspend fraudulent ad accounts before they could reach users.
AI Supports Human Experts in the Ad Safety Efforts
Google’s LLM-based fraud detection system allows it to proactively identify fraudulent activity before it occurs. The company rolled out 50 LLM upgrades to enhance security checks across its ecosystem. However, human experts still play a critical role in verifying AI-generated decisions. The anti-fraud team consists of 100 professionals, including members from the Ads Safety team, Trust and Safety division, and DeepMind researchers.
“While these AI models are very, very important to us and have delivered a series of impressive improvements, we still have humans involved throughout the process,” said Alex Rodriguez, a general manager at Google Ads Safety during a virtual media roundtable.
Google Analyzes Deepfake Ads Scam
One of the biggest security challenges of 2024 was the rise of deepfake advertising scams, which involved fake celebrity visuals. Google analyzed these scams to develop new technologies and policy frameworks to combat such illegal content. Over 700,000 company accounts distributing manipulated content received suspension notices. Google also reported a 90% reduction in the number of deepfake ad reports.
U.S. and India Saw the Biggest Action
The highest number of advertiser account removals occurred in the U.S. and India. In the United States, Google removed 39.2 million advertiser accounts and 1.8 billion advertisements. Key violations included ad network abuse, false healthcare claims, personalized ads, trademark infringement, and misrepresentation.
In India, 2.9 million accounts were suspended, with 247.4 million ads taken down. Violations mainly involved financial services, trademark misuse, ad network abuse, personalized ads, and gambling-related content. Globally, Google terminated 5 million scam-related accounts and 500 million scam advertisements.
Election Ads and Reduced Harm
Many countries conducted their elections in 2024. Google conducted a verification process which led to approving 8,900 new election advertisers. It also removed 10.7 million political ads, although political content represented only a small fraction of the total ad volume.
Overall, the number of advertisements restricted by Google was 9.1 billion. The count of publisher pages removed reached 1.3 billion.
Human Review and Policy Transparency Remain Crucial
Many advertisers expressed confusion over the suspension of their accounts. Google acknowledged that its alert messages were often unclear and committed to improving transparency in policy violation communications. Human reviewers have increased their intervention to ensure fair and accurate decision-making.
Google will continue to use AI to detect and suspend Fraud Ads Accounts in the future. The company asserts human supervision, and precise communication will continue to anchor its approach.
