Android adds optional AI scam checks to Google Messages

In short: Google is adding a new optional “enhanced protection” mode in Google Messages that uses AI to judge whether an unknown sender may be trying to scam you.

What happened

Google said it is rolling out a new setting in Google Messages called enhanced protection in the coming weeks. It builds on the app’s existing scam detection feature.

The new mode uses Gemini, Google’s AI system, to look at conversations with people who are not in your contacts. The goal is to spot “slow-burn” scams, where someone chats normally for a while before asking for money, gift cards, or personal information.

Google says this version uses cloud-based AI, meaning the checking happens on Google’s servers instead of only on your phone. Google also says it still uses on-device AI for some checks, including looking at unfamiliar links before you open them to see if they might be dangerous (like a “fake door” that leads to a phishing site).

Google says messages are encrypted during this AI analysis and processed in a secure, dedicated space that Google cannot access. The feature is opt-in, which means it is off by default and you choose whether to turn it on.

Google also announced a few other Android protections. These include blocking apps from turning on call forwarding and warning you if something tries to do it, plus hiding sensitive info during screen sharing in Chrome and Android WebView (a built-in tool many apps use to show web pages).

Why it matters

Text message scams are a common way people lose money and personal data. Extra warnings in your main messaging app can act like a second set of eyes, especially when a message looks friendly at first but is actually trying to trick you.

Source: The Verge AI