Google Rolls Out Gemini AI's Scam Detection Feature to Android's Messages App
Google Rolls Out Gemini AI's Scam Detection Feature to Android's Messages App
摘要: By
Isaiah Richard
Published: Oct 15 2026, 3:00 AM EDT
Share on Facebook
Share on Twitter
Share on LinkedIn
Share on Reddit
Share on Flipboard
正文
Isaiah Richard
Published: Oct 15 2026, 3:00 AM EDT
Share on Facebook
Share on Twitter
Share on LinkedIn
Share on Reddit
Share on Flipboard
Share on Pocket
Google Gemini Scam Detection
has now stepped up its defenses in a meaningful way, rolling out a wave of new protections across Android devices, including Gemini's scam detection feature, designed to stop fraud before it reaches users.
Mobile scams have grown far more sophisticated than the obvious phishing texts of years past, and millions of smartphone users are feeling the pressure.
Google Rolls Out Gemini Scam Detection to Messages App
Google has announced
a new opt-in enhanced protection mode coming to the Google Messages app that uses its most capable Gemini models to evaluate whether an incoming message is a scam attempt. The setting allows Scam Detection in Google Messages to analyze the full context of a conversation with an unknown sender, going beyond simple keyword detection to understand the intent behind what a potential scammer is saying.
When the enhanced protection mode is turned on, messages are analyzed through Google's Private AI Compute system, which encrypts the data and processes it in a secure dedicated space that is inaccessible to anyone else, including Google itself. This cloud-based analysis works alongside existing on-device detection to identify novel and sophisticated attacks that standard filters may not catch.
Gemini Nano, the on-device version of Google's AI model, will also be used to evaluate links sent in Messages before a user taps on them. Scammers frequently register brand new websites that have not yet been cataloged as malicious, and this on-device analysis is designed to catch those unfamiliar links by integrating real-time web signals to assess whether a link can be trusted.
As part of the same rollout, Google has also redesigned its scam warnings inside Messages to give users more context and additional time before tapping any link sen