Google Suspends 39M Ad Accounts for Fraud Using AI
Google's Ad Fraud Crackdown: 39.2 Million Accounts Suspended in 2024
Google announced on Wednesday the suspension of 39.2 million advertiser accounts during 2024. This figure represents a more than threefold increase compared to the previous year’s total, demonstrating a heightened effort to combat ad fraud.
The search engine giant is utilizing large language models (LLMs) and indicators like fraudulent payment information and instances of business impersonation. This allows for the proactive suspension of the majority of potentially fraudulent ad accounts before any advertisements are displayed.
Leveraging AI for Enhanced Safety
Throughout the past year, Google implemented over 50 enhancements powered by LLMs. These improvements are designed to bolster safety enforcement across all of its platforms.
Alex Rodriguez, a General Manager for Ads Safety at Google, emphasized the continued role of human oversight. He stated that while AI models are crucial and have yielded significant progress, human experts remain integral to the process.
Combating Deepfake Scams
A dedicated team of over 100 specialists from Google, including members of the Ads Safety team, the Trust and Safety division, and DeepMind researchers, collaborated to analyze and counter deepfake ad scams.
These scams often involve the unauthorized impersonation of public figures. The team developed effective countermeasures to address this growing threat.
The company introduced over 30 updates to its advertising and publisher policies, alongside technical defenses. These actions resulted in the suspension of over 700,000 accounts involved in fraudulent activity, leading to a reported 90% reduction in deepfake ad reports.
Regional Breakdown of Account Suspensions
In the United States, Google suspended 39.2 million advertiser accounts and removed 1.8 billion ads. Key violations included abuse of ad networks, trademark infringement, misleading healthcare claims, and misrepresentation.
India experienced the second-highest number of account suspensions, totaling 2.9 million. Google also removed 247.4 million ads in India, primarily related to financial services, trademark misuse, ad network abuse, personalized advertising, and gambling.
Focus on Scam-Related Violations
Google reported suspending 5 million advertiser accounts specifically for violations related to scams. This contributed to the removal of nearly half a billion scam-related advertisements overall.
Election Ad Monitoring
In a year where half the world’s population participated in elections, Google verified over 8,900 new election advertisers. A total of 10.7 million election-related ads were removed.
However, Rodriguez noted that the volume of election ads was relatively small compared to Google’s total ad inventory and wouldn't significantly impact overall safety metrics.
Overall Ad Blocking and Removal Statistics
Google blocked 5.1 billion ads and removed 1.3 billion pages throughout the past year. For comparison, in 2023, the company blocked over 5.5 billion ads and took action against 2.1 billion publisher pages.
The decrease in these numbers, according to Google, reflects improvements in preventative measures. Early detection and suspension of malicious accounts reduce the number of harmful ads that are created and disseminated.
Furthermore, 9.1 billion ads were restricted during the year.
Transparency and Appeal Processes
Google acknowledges concerns regarding the fairness of its rule enforcement. The company provides an appeal process, including human review, to ensure appropriate action is taken.
Rodriguez highlighted efforts to improve clarity and transparency in communication with advertisers. Policy updates and enhanced messaging aim to provide clearer explanations for account suspensions and restrictions.
These improvements are a key focus for the team, continuing into 2025.
