Google Play Protect will now use AI to warn you if an app is acting sketchy

Most Android users use their phones with Google Play Services on them, which gives them access to Google apps such as the Google Play Store and the accompanying Google Play Protect services. At Google I/O 2024, Google announced a key AI upgrade to Play Protect to improve fraud and abuse detection against sketchy apps.

Google Play Live Threat Detection is a new feature that leverages the device’s on-device AI capabilities to improve its security. On-device AI will analyze additional behavioral signals related to the use of sensitive permissions and interactions with other apps and services. If suspicious behavior is detected, Google Play Protect can send the app to Google for additional review. It can warn users or disable the app if malicious behavior is confirmed.

Previously, Google relied only on cloud-based methods to scan apps for suspicious behavior. However, certain apps displayed different behavior when running on a device in the wild than in cloud detection circumstances or in other more constrained environments. This change targets such scenarios, as the on-device AI model will observe the app’s actual behavior on the device.

Google says suspicious behavior is detected on the device in a “privacy-preserving” way through Private Compute Core. Later this year, this new Live Threat Detection feature will be deployed to devices from OPPO, HONOR, Lenovo, OnePlus, Nothing, Transsion, Sharp, and other manufacturers alongside Google Pixel devices.