Google Ads Using New AI Model To Catch Fraudulent Advertisers

Google has rolled out a new AI model, ALF, to better detect policy violations and malicious activity in Google Ads. This multimodal large foundation model analyzes text, images, video, and other factors to identify fraudulent advertisers with 99.8% precision. ALF processes large batches of advertisers to compare interactions and spot suspicious outliers. Although it has higher latency than previous models, it delivers substantial gains in fraud detection performance.