Suspicious messages, unfamiliar links and exposed login screens are targets of a new Android security rollout. Google detailed the changes on October 14, including an optional enhanced-protection setting in Google Messages.
The update matters because the protection operates at several points in a conversation. It can assess message context, warn before an unfamiliar link opens and hide sensitive fields during screen sharing.
Google says the enhanced setting adds analysis by cloud Gemini models for conversations with unknown senders. It uses Private AI Compute, a service the company describes as an encrypted, dedicated processing environment. Those privacy assurances are Google's description of its system.
On-device Gemini Nano also supports checks of unfamiliar links using current web signals. The company says the enhanced setting will roll out in the coming weeks, rather than being available immediately to every device.
Other changes extend the hiding of sensitive information during screen sharing to Chrome and Android WebView. Call-forwarding protections also add warnings and restrict background changes by apps.
Google reports that its existing phone and messaging tools flag more than 5 million suspected scam calls and texts each month. That measures flagged activity, not confirmed financial losses prevented.
The announcement also cites a comparative security assessment funded by Google. Its conclusions should be read with that funding relationship and tested device selection in mind.
Availability varies by device and feature. The release combines a staged rollout, existing on-device defenses and an additional opt-in setting.
