Your Pixel can now detect malicious apps that collect data without your consent

Google showcased a new Play Protect feature at I/O this May that utilized on-device AI models to protect users from sketchy apps. The Live Threat Detection feature is finally rolling out to Android devices, starting with Google’s Pixel lineup.

The new feature will analyze how apps interact with sensitive permissions and other apps and services to identify harmful apps. When it detects suspicious behavior, live threat detection will alert users in real time, prompting them to take immediate action to protect their devices. As you can see in the following image, the feature will deliver a notification when it finds an unsafe app with a link to Play Protect to help users remove the app from their devices.

Google says the feature can identify malicious apps that “try extra hard to hide their behavior or lie dormant for a time before engaging in suspicious activity.” Live threat detection will initially focus on apps that collect personal data without user consent, but Google plans to explore expanding its detection to other types of shady apps in the future.

Google has also assured that the feature won’t collect user data, with the on-device AI models analyzing app behavior in a privacy-preserving way through Private Compute Core. Live threat detection is now rolling out in Google Play Protect to Pixel 6 and newer models. Devices from other manufacturers will also receive the feature in the coming months.