Google has announced a new Android “Identity Check” security feature that locks sensitive settings behind biometric authentication when outside a trusted location. This feature is part of Android’s theft protection suite, which already includes comprehensive theft prevention mechanisms. Some sensitive actions protected by Identity Check include performing a factory reset, changing the screen lock, registering a new fingerprint, turning off ‘Find My Device’, adding a Google account, accessing developer options, and opening Google Password Manager.
At launch, Identity Check will be available exclusively on Google Pixel running Android 15 and Samsung Galaxy phones running One UI 7. To enable this feature, go to Google account settings under All Services, select Theft Protection, and toggle the “Identity Check” switch. Google is also expanding the availability of its AI-powered Theft Detection Lock.
This feature can identify phone theft events based on sensor data, Wi-Fi, and Bluetooth, subsequently locking the screen if someone grabs the phone and runs away. Initially only available on Google Pixel devices, Theft Detection Lock is now rolling out to all phones running Android 10 and later. To activate the Theft Detection Lock on your device, navigate to Google account settings under All Services, select Theft Protection, and toggle the “Theft Detection Lock” switch.
New Android security feature details
This feature can be introduced via Google Play system updates, even on devices no longer supported by regular security updates. Google has indicated it is working closely with the GSMA to develop advanced systems to tackle mobile device theft.
More information on these efforts will be disclosed at a future date. Since many thefts can be violent or involve sudden, jerky movements, Google sought advice from its Activity Recognition team, which worked on somewhat similar features. At a very high level, it starts with motion sensors,” says Jianing Sandra Guo, an Android security and privacy product manager.
Google developers researched smartphone theft methods globally and role-played the robberies for training data. “We reenact those scenarios. Then we gather the trace of the device,” says Guo.
“We use AI to compile those traces into a model to try to predict the next time when a snatch happens.” For Android owners, this process is done using on-device models. The AI model is not static, and developers plan to continue adding new smartphone theft scenarios to the training data as criminals modify their behavior.
