Google brings Gemini-powered scam alerts to Android
Google introduced new Android safeguards against scams and privacy intrusions. An opt-in Messages feature will use Gemini to judge whether incoming texts are deceptive and alert users. The rollout also adds on-device link scanning, clearer warnings, automatic hiding of sensitive data during screen sharing, and call-forwarding protection.
Google's Messages app will offer an opt-in Scam Detection setting that uses Gemini models to assess a sender's intent and flag potentially deceptive texts. Messages are processed in Private AI Compute, an encrypted environment the company says even it cannot access, while Gemini Nano checks links on-device. The feature reaches Pixel 10 and newer, Galaxy S26, and foldable Galaxy devices, among others.
Scam Detection already appeared in chat notifications on Pixel phones and is spreading to more hardware, with Honor, Motorola, OPPO, OnePlus, realme, vivo and Xiaomi expected to enable it later this year. Separately, Android 15+ devices gain automatic screen-sharing privacy this month, and call-forwarding protection arrives in December.
Android's scale means these safeguards could reach a vast number of people, potentially reducing losses from text-based fraud and impersonation calls. Opt-in design may limit how many users actually benefit, and encrypted on-device processing could ease privacy concerns. Banks and their customers in Brazil already see verification checks, hinting at how institutions might rely on platform-level defenses. Still, scammers adapt quickly, so the long-term effect may depend on how consistently these tools are deployed across manufacturers and regions.