Android adds Gemini scam detection to Messages for millions
On 14 October 2026 Google announced an opt-in Enhanced protection mode in Android Messages using cloud Gemini models, part of scam tools flagging over 5 million suspected scam calls and texts monthly.
Photo: EngadgetKey points
Google announced an opt-in Enhanced protection mode in Android Messages using cloud Gemini models, part of scam tools flagging over 5 million suspected scam calls and texts monthly.
Google announced on 14 October 2026 an opt-in Enhanced protection mode in Android Messages that uses cloud Gemini models to detect scam messages, part of a set of scam and privacy tools for Android. Google's Scam Detection tools now flag more than 5 million suspected scam calls and texts every month, according to the company's Android Security blog post authored by Alberto Pastor Nieto and Shaifali Bhandari.
The new mode matters because it extends scam detection beyond on-device analysis to cloud models that can evaluate a conversation's context and a sender's intent more quickly with unknown senders. Google said scammers often disguise their motives behind slow-burn conversations, and that the cloud analysis augments existing on-device detection to identify novel and sophisticated attacks that try to hide their tracks.
How the cloud analysis works
With Enhanced protection enabled, Scam Detection draws on Google's cloud Gemini models through Private AI Compute, which encrypts messages and processes them in a secure dedicated space inaccessible to anyone else, Google included. A second layer uses on-device Gemini Nano to check real-time web signals on unfamiliar links before the user taps, since attackers routinely register brand-new websites that existing protections have not catalogued.
A third layer redesigns warnings to give users extra context and time before tapping links from numbers outside their contacts, a checkpoint previously limited to select countries or devices with Android Advanced Protection. Users can optionally report a message as spam, sending its contents to Google for analysis. Enhanced protection mode is opt-in and rolls out over the coming weeks.
The Private AI Compute analysis works on Pixel 6 and newer, Galaxy S24 series and newer, Galaxy Z Flip and Z Fold6 and newer, and other Android AICore devices. The Gemini Nano link checks require Pixel 10 and later, the Galaxy S26 series, Galaxy Z Flip and Z Fold8, and other select devices. Redesigned link warnings work on devices running Android 6 or later.
Which devices get the feature
Google is also expanding Scam Detection in chat notifications, which flags suspicious messages in the notification shade before the user opens a chat app, beyond Pixel to manufacturers including HONOR, Infinix, iQOO, itel, JOVI, Motorola, OPPO, OnePlus, realme, TECNO, vivo, and Xiaomi. Google said the on-device protection will reach hundreds of millions of additional users later in 2026.
Automatic screen-sharing privacy that hides passwords, banking information, and credit card security numbers will extend to Chrome and Android WebView on devices running Android 15 or later, rolling out during October 2026. Call forwarding protection, blocking apps from silently enabling call forwarding and warning before a manually entered MMI forwarding code takes effect, arrives in early December 2026 on Pixel and other select devices.
Wider scam tools and dates
To counter bank impersonation, Android partnered with Banco do Brasil, Itaú Unibanco, and other banks on verified financial calls that check in real time whether an incoming call was genuinely initiated by the bank. Android has blocked over three million fake calls in Brazil since launching the feature in June, and Google said it will work with more partner banks in the future.
A Google-commissioned evaluation by Leviathan Security Group, published October 2026, compared the Pixel 11 Pro, iPhone 18 Pro, Galaxy S26, and Motorola Razr 2026 on built-in, no-cost features enabled by default, finding Android devices delivered the strongest default scam protections of the group. A Google-YouGov survey conducted June and July 2026 of 4,500 adult smartphone users in the US, Australia, and Canada found 40% described themselves as very or extremely concerned about mobile fraud.
The survey included 2,400 respondents in the US, 1,200 in Australia, and 900 in Canada, with comparative scam-experience questions answered by a qualified subset of 2,658 respondents. Google said iPhone users were significantly more likely than Pixel users to report falling victim to a scam and to call their device not effective at all at protection. Enhanced protection mode in Google Messages rolls out over the coming weeks. Call forwarding protection follows in early December 2026 on Pixel and other select devices, while screen-sharing privacy for Chrome and Android WebView arrives during October 2026 on Android 15 and later devices.
Frequently asked questions
What is Android's new Enhanced protection mode?
Enhanced protection mode is an opt-in setting in Android Messages announced 14 October 2026 that uses cloud Gemini models through Private AI Compute to evaluate message context and a sender's intent with unknown senders, supplementing on-device detection.
How many scam calls and texts does Google flag monthly?
Google's Scam Detection tools now flag more than 5 million suspected scam calls and texts every month, according to its Android Security blog post published 14 October 2026.
Which devices support the Gemini link checks?
On-device Gemini Nano link checks using real-time web signals are available on Pixel 10 and later, the Galaxy S26 series, Galaxy Z Flip and Z Fold8, and other select devices.


