Google Puts Gemini on Scam-Text Duty in Android Messages (Opt-In Enhanced Protection)
Google ● Covered by 3 sources
Google is adding Gemini-based scam checks to Android Messages, with an opt-in “enhanced protection” mode. It’s trying to catch smarter text scams before you tap, plus fake calls and screen-share tricks.
Based on reporting by Google — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Google is pushing a new layer of scam defenses into Android Messages and the broader phone stack, and this time the pitch is very explicitly AI. The company says mobile fraud has moved from clumsy robocalls to spoofed numbers, fake bank alerts, and voice cloning, and that 40% of smartphone users surveyed across the U.S., Canada, and Australia say they’re very or extremely concerned about it.
The headline change is an opt-in Enhanced protection mode in Google Messages. When it’s turned on, Scam Detection can tap Google’s cloud Gemini models through Private AI Compute to analyze messy, slow-burn conversations with unknown senders. Google says that setup is meant to catch novel scam tactics faster while keeping the data encrypted in a dedicated space that’s private to the user and inaccessible to anyone else, including Google.
There’s more on the device itself too. Google Messages will use Gemini Nano for smarter link checks, looking at unfamiliar links with real-time web signals before you tap. It’s also adding clearer warnings for links from numbers that aren’t in your contacts, a feature that had previously been limited to select countries or Android Advanced Protection devices. And scam alerts that already appear in Pixel chat notifications are being expanded to hundreds of millions more users through Android device makers later this year.
The call side is getting its own defenses. Android will now blank sensitive information on screen shares not just in native apps, but also in Chrome and Android WebView, and it’s blocking apps from quietly turning on call forwarding in the background. If someone tries to trick a user into entering a forwarding code manually, Android will warn before it takes effect.
Google is leaning hard on outside validation, too. A Google-commissioned Leviathan Security Group report says Android offers the strongest out-of-the-box scam protection versus iOS, and Google says its systems already flag more than 5 million suspected scam calls and texts each month. In Brazil, Android says it has blocked more than 3 million spoofed calls from Banco do Brasil and Itaú Unibanco since June, which tells you where the arms race is headed: the phone itself is becoming the bouncer.
My take — AI-written commentary, not fact-checked reporting
This is the right direction, because scam defense has become a platform problem, not a user problem. The annoying part is that the most useful protection now comes with a cloud-sized AI stack behind it, which is exactly how every “simple safety feature” ends up in 2026. Still, if Android can quietly kill fake bank calls and shady text traps before they land, that’s one of the rare pieces of AI hype that actually pays rent.
Read more about this at: Google
