Google is adding an optional layer of scam protection to Android. In Google Messages, the new enhanced protection mode will use Gemini to assess conversations with unrecognized senders, while other announced features target suspicious links, call forwarding and screen sharing.
Gemini will assess conversations with unknown senders
The new setting builds on the Scam Detection feature already available in Google Messages. It will use cloud-based Gemini AI models to examine conversations with people who are not recognized contacts and assess whether their intentions may be fraudulent.
That approach is meant to catch schemes that take time to unfold. Google says scammers can disguise their motives through slow conversations, so cloud analysis may help identify novel or more sophisticated attacks that existing on-device detection could miss.
Google describes the cloud analysis as secure and says user data is encrypted and “processed in a secure dedicated space” that Google and others cannot access. The feature is opt-in, giving users a choice about whether to enable this additional analysis.
Unfamiliar links get an on-device check
Enhanced protection will also use Gemini Nano, which runs on the device, to inspect real-time web signals associated with unfamiliar URLs received in messages. The check happens before a link is opened and is intended to help determine whether it may be malicious.
The two AI components therefore address different parts of a potential scam: the cloud-based Gemini models evaluate the conversation, while Gemini Nano checks links on the phone. Together, they add to the existing detection system rather than replacing it.
More Android safeguards cover calls and screens
Google also announced protections beyond messaging. Android will block apps from enabling call forwarding and notify the user when an attempt is made. That gives users a warning when an app tries to change how their calls are routed.
Another feature will automatically conceal sensitive information during screen sharing in Chrome and Android WebView. The measure is intended to keep private details from appearing to someone viewing a shared screen.
These changes address different ways a user might be exposed: a message can lead to a harmful link, a call-forwarding change can alter where calls go, and screen sharing can reveal information on display. Google’s announcement describes protections for each of these situations.
Chat scam detection is expanding across brands
The scam detection feature for chat notifications, introduced on Pixel phones last year, is expanding to devices from Honor, Motorola, Oppo, OnePlus, Realme, Tecno, Vivo, Xiaomi, and more. Google says it plans to bring the feature to “hundreds of millions of additional users later this year.”
The rollout means that these protections are not limited to Pixel devices, though the announcement does not specify which individual models will receive them. The enhanced protection setting in Google Messages is also described as rolling out in the coming weeks.
For users, the central choice is whether to turn on the optional message protection mode. Its combination of conversation analysis and link checks is designed to make suspicious messages easier to spot, while the call and screen protections add safeguards elsewhere in Android.