Google is adding an opt-in "enhanced protection" setting to Scam Detection in Google Messages that hands conversations with unrecognized senders to its larger cloud Gemini models to judge what the sender is really after, Engadget reports. It's aimed at the patient kind of scam, like Google's example of a "wrong number" text about a dog photo that turns into small talk and a request for your name. "Scammers often disguise their true motives behind slow-burn conversations," Google said, per The Verge.
The analysis runs in Private AI Compute, which Google says encrypts the messages and processes them somewhere even Google can't see. On-device Gemini Nano also checks links from unfamiliar senders before you open them. The setting rolls out over the coming weeks on the Pixel 10 and later, the Galaxy S26 series, the Galaxy Z Flip and Fold 8, and other select phones.
Google announced a few other pieces too. Scam warnings inside chat notifications, which started on Pixel, are coming to phones from Honor, Motorola, OPPO, OnePlus, realme, vivo and Xiaomi later this year. Starting later this month, Android 15 and newer will automatically hide passwords, banking details and card security codes when you share a screen showing Chrome or Android WebView. In early December, Pixel and select other phones will block apps from quietly turning on call forwarding and warn you if someone talks you into typing a forwarding code. In Brazil, a bank-call verification check has blocked more than three million fake calls since June, and there's no word on a U.S. version yet.
— via Engadget