Chrome Utilizes AI Against Tech Support Scams in New Update

by Justin Erickson

Google is adding an on-device layer of scam protection to Chrome 137. Powered by the Gemini Nano large-language model, the feature analyzes web pages locally for hallmarks of tech-support fraud—full-screen alerts, keyboard-lock API abuse, and scare-ware language—then consults Safe Browsing for a final verdict before a warning appears. Google says running the model on the user’s machine cuts response time and keeps page data private. The control will roll out first to desktop users who turn on Chrome’s Safe Browsing “Enhanced Protection” feature (Settings > Privacy & Security > Security > Enhanced Protection). Google also says that detection runs asynchronously to limit performance impact. The same Gemini-backed system may extend to fake package-tracking and toll-fee sites later this year, as well as Chrome for Android.