Google has launched two new AI safety tools to protect Android users. These tools monitor activity in real-time to protect against phone call-based fraud and dangerous apps. The first tool, "Scam Detection in Phone," examines conversation patterns to identify potential fraudulent calls. The second tool, "Google Play Protect Real-Time Alerts," monitors the background activity of the app after it is installed and detects suspicious apps.

Google said that these features will currently be available for Pixel 6 and newer models. The "Scam Detection in Phone" tool will first be available only to users in the Google Beta program in the US and will only work on English language calls. At the same time, Google Play Protect alerts will also be available outside the US.

This new scam detection feature differs from traditional caller IDs, which only track numbers and calling patterns. Instead, Google's machine learning model processes the call's conversation patterns in real-time to detect whether the call may be fraudulent. For example, if the caller claims to be from a bank and asks for a money transfer, this AI model can analyze his voice information to check if similar patterns have been used in scams before.

If the call seems likely to be a scam, the AI will provide an audio and haptic alert and also show a visual warning. This feature will be off by default and users can turn it on for all calls or only for a specific call. Google said that no audio or transcription is stored in this process, nor is it sent to Google servers.

The second feature is part of Google Play Protect, which identifies dangerous apps on the Play Store. This new "Live Threat Detection" feature is powered by AI, which monitors potentially dangerous apps in real time. If the activity of an app looks suspicious, this tool will warn the user in real time.